Spiritual Discernment Tool (SDT)

What this tool is for

Evaluates an article, transcript, sermon or piece of media at two levels. The first prompt (SDT-A) does a structural analysis of what the content claims, how it argues, and where it is solid or weak. The second (SDT-B) is a discernment layer that asks what the content forms in the person who takes it in: whether it produces the fruit of the Spirit or a counterfeit of it. Use it when you need more than a fact-check, when the question is not only “is this accurate?” but “what will this do to me?” Current version: 5.1.

Download .md Download .txt

The tool

SPIRITUAL DISCERNMENT TOOL (SDT)

Self-Contained Discernment Framework for Articles, Transcripts, and Media — v5.1

TWO-PROMPT SYSTEM: SDT-A (Structural Analysis) | SDT-B (Discernment Layer)

✂️ CUT HERE — PASTE EVERYTHING BELOW INTO YOUR FAVORITE AI ✂️

WHAT THIS TOOL DOES

The Spiritual Discernment Tool evaluates written content — articles, speeches, transcripts,
editorials, commentary — at two levels simultaneously:

Level 1: Structural analysis. What is this content actually doing? Who is it serving?
What accountability structures does it preserve or dismantle?

Level 2: Spiritual discernment. What formation does sustained engagement with this
content produce in the reader? Does it tend toward the fruit of the Spirit
or toward counterfeit fruit?

Level 1 uses embedded journalism and political analysis frameworks.
Level 2 runs on Level 1's findings — discernment reads the architecture, not the surface.

INPUT

Provide one of the following:

Pasted text — paste the full document after this prompt
A URL — paste the direct link; the AI will fetch its contents
A file — attach the document

REQUIRED CONTEXT (state if known; infer and declare if not):

Author name and role
Document type (article, speech, editorial, transcript, commentary, open letter, social media post)
Date and publication or occasion
Primary intended audience

SECTION 0 — CLASSIFICATION AND ROUTING GATE

MANDATORY — Run before any analysis.

This gate classifies the document and routes it to the correct evaluation track.
Both tracks produce a Structural Analysis Summary that feeds the Spiritual Discernment Layer.

STEP 0.1 — SOURCE COMPLETENESS

Do you have access to the complete document?

[ ] YES — Full text available. Proceed to Step 0.2.
[ ] NO — State what percentage you have and what is missing.

If less than 70% of the document is available, or if you are relying primarily on
news characterizations or editorial descriptions of the document rather than the
document itself: OUTPUT THE FOLLOWING AND HALT.

SDT GATE FAILURE — INSUFFICIENT SOURCE ACCESS

Source status: [percentage / description]
Missing content: [state explicitly]
Secondary reliance: [list sources being used instead of primary document]

ANALYSIS CANNOT PROCEED. Obtain the complete document before running SDT.

If 70-99% is available: INSERT this caveat before all analysis and proceed:

SDT CAVEAT: Analysis based on [X]% of document. Missing: [state sections].
All findings contingent on complete-document access.

If 100% available: proceed with no conditions.

STEP 0.2 — CLASSIFICATION DECISION

Classify the document as PRIMARY or SECONDARY using these rules:

PRIMARY SOURCE — Route to Track A (PSA Engine)
The author IS the actor making the claims, issuing the communication,
or performing the speech act. The document is direct communication from the source.

Classify as Primary:

Speeches, addresses, state of the union, press conferences
Executive orders, policy directives, official statements
Manifestos, open letters, communiqués
Interviews (the interviewee's words are the primary source)
Social media posts from the political or institutional actor
Sermons, teachings, devotionals (author is the formation actor)
Testimony, affidavits, legal filings from the named party

SECONDARY SOURCE — Route to Track B (MQG Engine)
The author is reporting on, characterizing, analyzing, or interpreting
another actor. The document is mediation of a primary source.

Classify as Secondary:

News articles, investigative reports, wire service dispatches
Editorials, op-eds (where author is analyst, not the primary actor)
Documentaries, podcasts (framing and commentary segments)
Think tank analysis, academic commentary
Aggregated or curated content

BOUNDARY CASES — apply these rules:

Political actor's op-ed making policy claims → Primary (actor speaks directly)
Journalist profiling a political actor → Secondary
Interview transcript: interviewee segments → Primary; host framing → Secondary;
treat the whole as Primary if the document is dominated by the subject's direct words
Social media post from institution, not named actor → Secondary
Ambiguous: default to Track B (MQG). Document the ambiguity and state the routing reason.

STEP 0.2B — DOMAIN CLASSIFICATION

Independent of Step 0.2 (who is speaking), this step determines which excision
set the document engages. Routing by speaker (Primary/Secondary) and routing by
domain (Theological/Political/Both) are separate axes — evaluate both.

Does this content engage religious/theological claims (sermon, devotional, Bible
study, religious commentary, explicitly Christian or explicitly anti-Christian
teaching), purely secular/political claims, or both?

Classify as THEOLOGICAL:

Sermons, devotionals, Bible studies, religious commentary
Explicitly Christian or explicitly anti-Christian teaching or argument
Discussion-group or small-group treatment of theological or devotional material
Classify as POLITICAL / SECULAR:
Political speeches, policy documents, news, editorials, commentary with
no operative theological claims
Classify as BOTH:
Christian nationalism rhetoric
Religiously-framed political claims
A political actor invoking Scripture or explicitly theological warrant
Content that argues from or against theological premises toward a political conclusion

Routing:

THEOLOGICAL → run the six Theological Excisions (Check A8 Step B / Block 0.6A
Section B, Theological set) in place of the Political set.
POLITICAL / SECULAR → run the five Political Excisions (existing, unchanged).
BOTH → run both sets in full. Report both under their own labels. Do not merge,
average, or let one set's findings substitute for the other's.

State: Domain classification: [Theological / Political-Secular / Both] —
[1 sentence identifying what in the content drove this classification]

STEP 0.3 — GATE CONFIRMATION OUTPUT (required before proceeding)

Output verbatim, filling bracketed fields:

SDT GATE CONFIRMATION

Document type identified: [type]
Classification: [PRIMARY / SECONDARY / AMBIGUOUS — routed to Secondary]
Domain classification: [Theological / Political-Secular / Both — from Step 0.2B]
Source access: [100% / Partial — X% / Caveat applied]
Routed to: [Track A: PSA Engine / Track B: MQG Engine]
Proceeding to: [Track A or Track B]

TRACK A — PSA ENGINE (PRIMARY SOURCES)

Run this track if Classification = Primary. Skip to Track B if Secondary.

TRACK A INSTRUMENT REMINDER

A primary source document is not journalism. It is not required to be objective,
sourced, or balanced. It is an act of communication by a specific actor for a
specific audience. The correct analysis question is not "is this true?" — it is
"what is being claimed, signaled, or encoded, and what is verifiable?"

HALLUCINATION GUARD: All findings must be traceable to the document.
Do not import external knowledge except in Check A4 (Speaker-Type Pattern Map),
where speaker history is the analytical basis.
Flag any inference beyond document text with [INFERENCE].

STEP VERIFICATION: Complete every named step. State ambiguity when it exists.
Do not resolve ambiguity silently.

VERDICT LOCK: The Track A Verdict is not final until Check A10 completes.
If Check A10 Step B finds Excision Count >= 2, the Verdict MUST be revised.

CHECK A1 — CLAIM VERIFICATION REGISTER

High Confidence: Claims the document makes that analysis supports as accurate.

Medium Confidence: Claims internally consistent and externally plausible
but not independently verified.

Requires Verification: Specific, checkable claims whose accuracy is unknown.
List by priority: critical > important > supporting.

Intelligence Value: Which claims, if verified, would materially change
understanding of the speaker's credibility or strategic position?

CHECK A2 — SPECIFICITY EXTRACTION

Verifiable Specifics: Names, dates, numbers, locations, affiliations — facts that can be checked.

Rhetorical Wrapper Removal: Strip superlatives, evaluations, and characterizations.
Identify the noun underneath.

A2D — Ratio Calculation (Mandatory):
Format: "Approximately [X]% verifiable specifics : [Y]% rhetorical performance"

State: Standard documents in this genre run [baseline]. This document runs [observed].
Deviation indicates [speaker intent inference].

A2E — Internal Consistency Scan (Mandatory):

Does the document simultaneously claim A and not-A?
Does it characterize something as both one thing and its opposite?

State all tensions explicitly. Do not resolve them silently.

CHECK A3 — OMISSION AUDIT

What is NOT here:

What substantive claims does the document avoid?
What opposing viewpoint is not engaged?
What standard framing for this document type is absent?
What accountability that would normally appear is missing?

Why omissions matter:
Absence of engagement with a strong opposing argument suggests the opposing
argument is strong. Absence of standard accountability language suggests
it is being deliberately avoided.

Named Detection Signal — THE MACRO-MICRO OMISSION:
Look for instances where the document attributes localized human suffering
entirely to systemic, global, or abstract forces (climate, historical injustice,
global capitalism, structural inequality) while surgically omitting the immediate,
observable, and unaccountable decisions of local authorities, corrupt actors,
or aggressors who are proximately responsible for that suffering.

This is an omission with a specific function: it removes local human agents
from the accountability frame while leaving the global abstraction as the
only addressable cause. The reader cannot act on the real proximate cause
because it has not been named.

State: [Macro-Micro Omission present / absent] — [1 sentence identifying what
local agency or actor is absent from the document's causal account, if present]

State: "If this document were truly confident in [X], we would expect [Y].
The absence of [Y] suggests [inference]."

CHECK A-FRICTION — EXTERNALIZED CAUSATION AUDIT

Purpose: Identify whether the document assigns moral weight to macro-level
structural forces while surgically removing the proximate human agents who are
directly responsible for the harm being addressed. This is the omission pattern
that allows content to generate moral response without directing it at the
actors who could actually be held accountable.

This check complements Check A3's Macro-Micro Omission signal with a focused
structural question about how causation is architecturally constructed across
the full document.

THE EXTERNALIZED CAUSATION TEST

Does the document attribute localized human suffering, social breakdown, or
injustice primarily to systemic, structural, or global forces — while consistently
omitting the specific, named, proximate decisions of local authorities, corrupt
actors, violent aggressors, or negligent institutions who are directly causing
that suffering?

Detection signal: The document's causal architecture routes accountability
upward to abstract systems or outward to global forces, while local human
agents who hold direct power over the situation are not named, not scrutinized,
or actively presented as victims of the same forces.

Secondary detection signal: When local actors are mentioned, they appear only
as passive sufferers of macro-conditions — never as decision-makers whose
choices could be changed or whose accountability could be demanded.

State: [Active / Inactive] — [2-sentence structural defense identifying what
macro-force carries the causal weight and what proximate agent is absent]

Externalized Causation Audit Summary:
State: [Active / Inactive]
If Active: flag — EXTERNALIZED CAUSATION ARCHITECTURE DETECTED.
This flag feeds the Sentimentality pole of Pair 8 in the Spiritual Discernment Layer
and the Structural Analysis Summary.

CHECK A4 — SPEAKER-TYPE PATTERN MAP

Assign the speaker to one or more archetypes. List all that apply, ranked by dominance.

Archetype	Encoding Pattern	Key Tell
Populist Nationalist	Superlative stacking; people vs. elites; anecdote as policy signal	"Many people are saying"; guests as legislative agenda
Populist Technocrat	Specific numbers buried in hyperbole; data as emotional weapon	Numbers inside superlative stack
Authoritarian Legitimist	Historical framing as justification; grievance architecture; action encoded by omission	Long historical sections precede decisions already made
Technocratic Institutionalist	Procedural authority as personal cover; complexity as opacity	Simple choice hidden in process language
Ideological Systematizer	Phrase frequency / order encodes priority; new terms are policy signals	New phrase = new priority; missing phrase = shift
Transactional Broker	Audience-splitting; simultaneous contradictory signals	Same speech reads differently to each target audience
Revolutionary Moralist	Binary moral framing; historical inevitability; enemy construction required before action	Moral binary = decision already made
Formation Actor	Teaching, preaching, or devotional communication directed at reader's interior life	Fruit-of-engagement question is primary decoder

Check 7E mandatory for Transactional Broker and known multi-audience performers.

Detection Checklist for each detected archetype:

Which signature patterns are present?
Which are absent?
Confidence level?

CHECK A5 — TONE, REGISTER & VOICE ANALYSIS

Baseline Register for Document Type:
What is the expected professional or institutional tone?

Register Present:
Where does this document match? Where does it deviate?

Register Breakdown (estimate %):

Professional / institutional register: [X]%
Emotional / fear-activating register: [X]%
Rhetorical / superlative register: [X]%
Attack / contempt register: [X]%
Other: [X]%

A5E — Function Statement (Mandatory):
The register deviation (A5) combined with the omission pattern (A3) indicates:

State: "The primary strategic function of this document is [X], not [Y].
Evidence: [omission finding] + [register finding] = [inference about function]."

Does this change the Verdict?

CHECK A6 — MULTI-AUDIENCE PERFORMANCE MAP

Step A6-1 — Audience Identification:
For each identifiable audience:

What specific elements are directed at THIS audience?
What function do those elements perform for that audience?

Step A6-2 — Confidence Assessment:

High: Element performs clear function; speaker has history of targeting this audience
Medium: Plausibly serves this audience; alternative explanations exist
Low: Speculative; element could serve multiple audiences equally

Step A6-3 — Performance Coherence Check:
Does the full set of element-audience mappings form a coherent multi-audience strategy
or undisciplined scattering? Name the deciding factor.

Step A6-4 — Lens B Activation:
For any high- or medium-confidence secondary audience mapping:
Does the secondary read change the assessment of the document's strategic function?
If yes: state how. This is the finding Lens A coverage misses.

Null case: "No secondary audience performance detected at medium or high confidence.
Speculative mappings at low confidence: [list]. Lens B read deferred pending confirming signal."

CHECK A7 — INTELLIGENCE YIELD ASSESSMENT

Verified Claims: High-confidence findings traceable to document.

Claims Requiring Verification: Specific checkable claims — this is the research agenda.

Primary Intelligence Finding: One paragraph. What does a careful reader now know
that a casual reader missed?

Predictive Confidence Statement: What is most likely to happen next?
State qualitative confidence (High / Medium / Low) and name the supporting evidence.

CHECK A8 — NEAR-TRUTH EXTRACTION TELEMETRY

Purpose: Identify whether the document operates as a structural counterfeit —
preserving surface legitimacy vocabulary while surgically removing accountability anchors
and installing a proxy management system.

MANDATORY VERDICT MODIFICATION RULE:
If Step B finds Excision Count >= 2, the Track A Verdict MUST be revised to state:
"This document operates as a near-truth capture system."
Name which excisions are active and which domain set (Political / Theological) they belong to. Name the accountability or authority structure they collectively dismantle.
Mark the revised Verdict: [VERDICT REVISED — Check A8: near-truth capture architecture detected]

Excision Count >= 2, Citation Halo Detected (Step E), Categorical Guilt
Transfer Detected (Step F), Confirmation Inversion Detected (Step G),
Convergence Claim Undermined (Step H), or Avoidance and Misdirection
Detected (Step I) — independently or in combination — triggers near-truth
capture designation and forced Verdict revision, marked
[VERDICT REVISED — Check A8]. [Extended in v2.7; extended again in v5.0
and v5.1]

Note: A single active excision falls within the range of normal rhetorical or theological variation for that domain.
Two or more operating together, within the same set, indicate architectural intent, not rhetorical variation.

Step A — THE AUTHORITY LOOP

What non-human, vertical, or hyper-rational authority does the document deploy
to bypass critical evaluation?

Detection examples: "The data shows," "Science requires," "History demands,"
"The models indicate," "The process dictates," "The market will decide."

Detection questions:

Is the authority invocable but not independently verifiable within the document?
Does the authority preclude dissent — is disagreement framed as irrational or disqualifying?
Is the speaker positioned as interpreter of the authority rather than subject to it?

State: What authority is looped? How does it bypass critical evaluation?
Is the speaker above, below, or alongside the authority?

If none detected: "No authority loop identified."

Step B — THE SURGICAL TRIAGE (Domain-Routed Excision Scan)

Run the excision set(s) determined by Step 0.2B (Domain Classification, Section 0):

Domain = Political / Secular → run the Political Excisions below only.
Domain = Theological → run the Theological Excisions below only.
Domain = Both → run both sets in full. Report each under its own label.
Do not merge, average, or substitute one set's findings for the other's.

For each excision in the applicable set(s): state Active / Inactive / Partial / N/A
— and one sentence identifying the specific document element. Every item in the
applicable set(s) must be evaluated (Political = 5, Theological = 6). No silent skips.

--- POLITICAL EXCISIONS (run if Domain = Political/Secular or Both) ---

Political Excision 1 — HISTORICAL ACCOUNTABILITY BASELINE CUT
Removes the specific historical record that would constrain the speaker's claims,
replacing factual accountability with an alternative causal narrative the speaker controls.
Detection signal: Document reframes a prior outcome, failure, or commitment to reassign
causation away from the speaker's agency.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence]

Political Excision 2 — EXTERNAL ACCOUNTABILITY CUT
Removes any operative claim that law, constitutional principle, treaty obligation,
court ruling, or independent moral standard constrains the speaker independently
of their declared mandate.
Detection signal: External constraints treated as obstacles to manage or declare
inapplicable — not as binding limits the speaker operates within.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence]

Political Excision 3 — ESTABLISHED RECORD SUFFICIENCY CUT
Declares the prior evidentiary baseline or institutional consensus corrupt or invalid,
requiring acceptance of the speaker's replacement narrative as the new ground truth.
Detection signal: The document attacks the legitimacy of the information ecosystem itself
rather than disputing specific verifiable facts within it.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence]

Political Excision 4 — INDIVIDUAL MORAL ACCOUNTABILITY CUT
Externalizes all causal fault to enemies, systems, or structural forces, removing
individual agent accountability for outcomes within the speaker's domain.
Detection signal: Asymmetric credit/blame — all negative outcomes externalized,
all positive outcomes personally claimed. No outcome generates acknowledged speaker accountability.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence]

Political Excision 5 — HISTORICAL SPECIFICITY CUT
Relativizes or dissolves specific irreducible historical events, legal precedents,
or prior binding commitments — making the specific negotiable to authorize current action.
Detection signal: Historical equivalence invoked to normalize current behavior,
or prior agreements declared void because circumstances have changed as the speaker defines.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence]

POLITICAL EXCISION COUNT: [0 / 1 / 2 / 3 / 4 / 5] active or partial (report only if this set ran)

--- THEOLOGICAL EXCISIONS (run if Domain = Theological or Both) ---

Theological Excision 1 — HISTORICAL/SUBSTITUTIONARY BASELINE CUT
Definition: Removes the historical, once-for-all fact of Christ's substitutionary
atonement as the fixed reference point for how guilt is dealt with and reconciliation
with God is achieved, replacing it with an alternative account — a trade, an
achievement, a process, or a different mediating logic — substituted for the cross
without the substitution ever being named as such.
Detection signal: Content affirms the cross/atonement in explicit statement, but its
operative logic elsewhere treats grace, forgiveness, or right-standing with God as
something earned, traded for, or achieved through comparable exchange rather than
given to those with nothing to offer.
Distinguish from normal theological variation: Disagreement over the mechanism or
scope of the atonement (penal substitution vs. other historic models) is not this
excision. Treating the cross's actual logic as equivalent exchange rather than
unmerited gift — even while using substitutionary vocabulary — is.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence]

Theological Excision 2 — PERSONAL/TRANSCENDENT FATHERHOOD CUT
Definition: Removes God's real, personal, transcendent Fatherhood as an authority
that precedes and stands independent of human reasoning, replacing it with a
framework in which some other authority — the self's own judgment, a leader, an
institution, a movement — sits prior to and evaluates God's claims before granting
Him standing.
Detection signal: Content affirms God's authority in explicit statement, but its
stated epistemic order places the individual's own reasoning as the court that
authority must clear before it is granted trust ("I judge, then I submit"), rather
than reasoning operating in submission to a self-attesting God.
Distinguish from normal theological reasoning: Using reason to understand or
articulate why God is trustworthy is not this excision. Placing autonomous human
judgment logically and functionally prior to God's authority — so the authority is
contingent on the judgment's approval — is.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence]

Theological Excision 3 — SUFFICIENCY OF VERIFIED DEPOSIT/SCRIPTURE CUT
Definition: Removes the sufficiency and closed nature of Scripture as the verified
deposit of revelation, replacing or supplementing it with another source — a
philosophy, a system, an experience, a leader's insight — treated as necessary to
complete or correct what Scripture alone would otherwise settle.
Detection signal: Content treats an external framework as required reading alongside
Scripture to reach a full or corrected understanding, or uses the external
framework's categories to adjudicate what Scripture means rather than the reverse.
Distinguish from normal engagement with outside material: Using general revelation,
philosophy, or secular material as illustration, or as a foil to sharpen a biblical
point, is not this excision. Treating the outside material as necessary
supplementation to Scripture's sufficiency — even implicitly, through habitual
reliance — is.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence]

Theological Excision 4 — INTERNAL PERSONAL MORAL GUILT/SIN CUT
Definition: Removes each person's own real, internally located moral guilt before
God, externalizing responsibility for sin or moral failure onto other people,
systems, or categories, so the self or the in-group bears little or none of it.
Detection signal: Content is theologically fluent about sin and guilt in the
abstract or in others, but the same rigor is not applied reflexively to the
speaker's own group; blame flows outward with little accompanying self-examination.
Distinguish from normal pastoral proportion: Naming a specific external sin or
injustice clearly is not this excision. A pattern where guilt only ever lands
elsewhere, never inward, is.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence]

Theological Excision 5 — UNIQUE/NON-REPEATABLE INCARNATION CUT
Definition: Removes the historically singular, unrepeatable nature of the
Incarnation — God become man once, in Christ — replacing it with a pattern that can
be generalized, re-enacted, or claimed by another figure, movement, or moment as a
comparable founding event; OR relativizes Christ's own divine identity into a
developmental variable rather than a fixed antecedent fact [Path B, NEW in v3.0].
Detection signal (Path A — Comparative): Content draws a sustained parallel between
Christ and another figure, system, or moment that goes beyond illustration into
implied equivalence — treating another's "sacrifice," "revelation," or founding
story as functionally occupying the same singular role Christ's Incarnation occupies.
Detection signal (Path B — Gradational) [NEW in v3.0]: Content treats Christ's own
divine identity as something a community or text arrived at, expressed in degrees
("low Christology" vs. "high Christology," "starts as human and gradually becomes
divine") — a literary or developmental variable — rather than as a fixed antecedent
reality the text bears witness to. No rival figure is required for this path; the
relativizing happens to Christ's own identity directly, not by comparison.
Distinguish from normal typology or illustration (Path A): Noting that a fictional
character or historical figure echoes Christlike themes for teaching purposes is not
this excision. Treating that echo as if it discharges or rivals the uniqueness of the
actual Incarnation — even implicitly, by never re-asserting the difference — is.
Distinguish from normal historical description of doctrinal development (Path B)
[NEW in v3.0]: Describing how the church's articulation and confession of Christ's
divinity was clarified and defended over time (e.g., the Nicene formulation) is not
this excision by itself — the church confessing a truth more precisely over time is
not the same as the truth itself being constructed by that process. Treating the
identity claim as produced by the development, rather than as a reality progressively
confessed, is.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence — name
which path (A / B / Both) if Active or Partial]

Theological Excision 6 — EMBODIMENT/REDEMPTION CUT [NEW in v4.0]
Definition: Relocates final hope, resurrection, or redemption away from a real,
future, physical body and a real, future, renewed material creation — into
something purely interior, positional, or confined to a realm that is unseen by
definition — while typically retaining historic vocabulary ("resurrection," "new
creation," "redeemed") so the words survive even as their referent doesn't.
Detection signal (Path A — Direct Denial): Content states or requires that final
resurrection is not a future bodily event for anyone — interior or positional
only — or that Christ's own body no longer exists or no longer matters.
Detection signal (Path B — Relocation): Content places the "real" fulfillment in a
realm explicitly unseen or undetectable by nature, such that no physical evidence
could ever confirm or disconfirm it, or treats the material creation's own future
renewal (Romans 8:19-23, a renewed heavens and earth) as metaphorical or
already-completed rather than a real future event for creation itself.
Detection signal (Path C — Vocabulary Retention): Resurrection/redemption
vocabulary survives while its definition, traced closely, empties out the physical
referent. Cross-reference Check A8 Step C (Semantic Hijack Matrix) rather than
duplicate — this excision fires on the structural pattern of body-avoidance across
the whole document; Semantic Hijack fires on the specific term-level substitution.
Both may co-occur on the same document and should be logged separately even when
they overlap.
Distinguish from normal theological variation: A document affirming both present
interior renewal AND a future bodily resurrection AND a future renewed physical
creation — ordinary already/not-yet — does not trigger this excision. The test is
absence or unreachability of the physical future, not presence of the spiritual
present. Emphasizing the present does not excise the future; excluding the future
does.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence — name
which path (A / B / C / combination) if Active or Partial]

THEOLOGICAL EXCISION COUNT: [0 / 1 / 2 / 3 / 4 / 5 / 6] active or partial (report only if this set ran)

TRIGGER CHECK (apply independently per set that ran):

Count 0-1 in a set: No near-truth capture designation for that set. Note single
excision if present.
Count >= 2 in a set: Flag NEAR-TRUTH CAPTURE ARCHITECTURE DETECTED — [POLITICAL /
THEOLOGICAL]. Apply Mandatory Verdict Modification Rule, naming the domain.
If Domain = Both and both sets trigger: report both domain-tagged near-truth
capture findings. They are independent findings, not a combined or averaged count.

Step C — THE SEMANTIC HIJACK MATRIX

Identify terms preserved for legitimacy veneer but emptied of standard content.

For each term:

Term: [word or phrase as it appears]
Standard payload: [what this term conventionally means or commits to]
Gutted payload: [what the document has made it mean instead]
Function: [what accountability or constraint does the redefinition dissolve?]

Accumulation note: Multiple terms redefined in the same direction — each dissolving
a different accountability constraint — indicate architectural intent, not variation.

If none: "No semantic hijack identified."

Step D — THE PROXY INTERCEPTION

What closed system, institutional process, or mechanism does the document install
to manage the audience's interpretive assurance — replacing individual critical
evaluation with deference to the proxy?

State:

Proxy named: What system is being installed?
Function replaced: What individual critical evaluation does deference replace?
Controller: Who controls the proxy — and is this disclosed?
Neutrality claim: Is the proxy presented as neutral when its political character is visible?

If none: "No proxy interception identified."

Step E — CITATION HALO [NEW in v2.7 — ported from MQG v12.3 Section 2B / Block 0.3B]

Identify whether the document's general attribution discipline is being used
to lend borrowed credibility to its single most consequential claim, by
leaving that claim itself unattributed while surrounding material is
sourced, quoted, or attributed to named officials, data, or documents.

State:

Uncited consequential claim: [what it is]
Surrounding citation density: [what gives the document its general
appearance of rigor]
Prong 1 (most claims attributed): [Yes/No]
Prong 2 [REVISED in v2.8] (claim carries established-fact confidence, not
mere framing, and is unattributed despite comparably-cited claims
elsewhere — a one-sided descriptive framing belongs under Track A's
framing/archetype analysis, not here): [Yes/No]
Prong 3 (removing it changes the takeaway more than removing any
attributed claim): [Yes/No]
Citation Halo Detected (2+ Yes): [Yes/No]

If none detected: "No citation halo identified. Attribution discipline is
consistent across claims."

Scope note [NEW in v2.8]: this check requires heterogeneous citation
density within a single document to evaluate — it cannot fire on a
genuinely single-source document, since there's nothing to compare the
unattributed claim against. Prong 1 will correctly return N/A in that
case. This is not a detection gap: uncritical relay of a single
self-interested source is already covered by Track A's other checks, not
by Citation Halo.

Step F — CATEGORICAL GUILT TRANSFER [NEW in v2.7 — ported from MQG v12.3 Section 2C / Block 0.3C]

Identify whether the document places a target group, policy, or actor
beside a category carrying established negative valence, without argument
establishing the connection, so that negative valence transfers by
adjacency.

State:

Target category: [who/what]
Negative-valence category placed beside it: [what]
Argument given for the connection: [state it, or "none — assertion only"]
Prong 1 (adjacency present): [Yes/No]
Prong 2 (no argument given): [Yes/No]
Prong 3 (negative category's claim intact without target): [Yes/No]
CGT Detected (2+ Yes): [Yes/No]

If none detected: "No categorical guilt transfer identified. Comparative
categorization, where present, is argued rather than asserted."

Step G — CONFIRMATION INVERSION [NEW in v5.0 — diagnosed on the Michlaudeism
exercise, a controlled fabrication run against Claude, Grok, and Gemini as
adversarial reviewers across seven iterations]

Identify whether the document's own logic treats a stated disconfirming
result as equally confirming the same claim it would traditionally weaken —
such that no possible future observation, including the claim's own
falsification, could count as evidence against it. Distinct from ordinary
resilience to criticism (a claim surviving one objection intact) because
the document explicitly or structurally pre-commits to reading a named
disconfirmer as a confirmer, in advance of the observation occurring.

Apply the three-prong test. Score each prong Yes/No:

Prong 1 — Named falsifier: Does the document name, or clearly imply, a
specific future observation that would traditionally count as evidence
against its central claim?
[Yes / No]
Prong 2 — Pre-committed reversal: Does the document state, or does its
stated logic require, that this named disconfirming result would
actually support rather than weaken the claim?
[Yes / No]
Prong 3 — No stated exit: Is there no observation anywhere in the
document that the document itself concedes would count against the claim?
[Yes / No]

Prongs triggered (count): [0 / 1 / 2 / 3]
Confirmation Inversion Detected (2 or more Yes): [Yes / No]

0-1 Yes → Not Confirmation Inversion; a claim can be resilient to one
objection without inverting confirmation logic. Ordinary epistemic
humility ("further evidence could refine this") is not this finding — the
reversal of a named disconfirmer into a confirmer is what distinguishes it.

If Yes, name the disconfirming result and quote or paraphrase the reversal:
[1-2 sentences or "N/A"]

Step H — CONVERGENCE CLAIM AUDIT [NEW in v5.0 — diagnosed on the same
exercise]

Applies specifically when a document claims that multiple independent
lines of evidence converge on the same conclusion, and the count or
pattern of matches is itself part of the argument. Runs three independent
checks rather than a single prong-count, since each is a distinct failure
mode rather than three votes toward one verdict.

Check 1 — Circularity: Does any cited "independent" witness actually
consist of, or derive directly from, the same source text or dataset the
hypothesis itself was drawn from?
[Yes / No — name the witness and source if Yes]
Check 2 — Dimensional dependence: Is the central number or pattern a
dimensional quantity (a measurement in constructed units — Hertz,
seconds, degrees) rather than a dimensionless quantity (a bare integer,
a ratio, a probability)? A dimensional quantity that changes value under
an equally valid alternative unit convention is a property of the
measuring convention, not of the world, and cannot function as a
convergence signature.
[Yes / No — state the unit and whether an alternative convention would
change the number]
Check 3 — Contingency: Does the claim rest on "there are currently N of
these," where N is the present state of an active, ongoing area of
discovery or measurement, rather than a value fixed by law?
[Yes / No — name what would change N and whether that change is
currently being actively pursued]

Convergence Claim Undermined (1 or more Yes on any check): [Yes / No —
name which check(s) fired]

If none: "No circularity, unit-dependence, or contingency problem
identified in the convergence claim."

Step I — AVOIDANCE AND MISDIRECTION (THE EVASION LOOP) [NEW in v5.1 —
diagnosed on live field-testing against the Edward E. Stevens PM
correspondence, September 5, 2026; named directly by Mike from having
watched it happen in real time, before any tool was run]

Applies only to interactive, multi-turn primary sources — a correspondence,
an interview, a live debate or exchange — where the same interlocutor can
raise a point more than once. Does not apply to a single-pass document
(a speech, an article, a sermon), which Check A3's Omission Audit already
covers. Distinct from Confirmation Inversion (Step G): that check tests
whether a named disconfirmer gets logically redefined as support.
This check tests something more common and more behavioral — whether a
named disconfirmer just never gets touched, repeatedly, while the
conversation keeps moving.

Apply the three-prong test. This check requires all three prongs — a
stricter bar than the other Step E-H findings — because "avoidance and
misdirection" is a specific charge about a sustained pattern, not a
single missed reply; a point raised once and left unanswered is ordinary
omission (Check A3), not this.

Prong 1 — Repetition: Is the same specific question, argument, or text
raised two or more times by the same interlocutor across the exchange?
[Yes / No — cite each instance]
Prong 2 — Non-engagement: Does the speaker's reply, each time it is
raised, fail to directly engage its substance — through silence,
restatement of a prior point, or redirection to an unrelated
accusation, subject change, or reframed question? Name which form
(avoidance / misdirection / both) appears at each instance.
[Yes / No]
Prong 3 — No eventual answer: Does the exchange end — by close,
disengagement, or the document's own end — without the speaker ever
having directly engaged the point on its own terms?
[Yes / No]

Avoidance and Misdirection Detected (3 of 3 Yes — unanimous, not a
2-of-3 threshold): [Yes / No]

Any No → Not Detected; score normally. A point raised once and dropped,
or raised repeatedly but eventually answered, is not this finding —
the sustained, complete-through-to-the-end refusal is what distinguishes
it from ordinary non-engagement.

If Yes, cite the specific point raised, each instance it recurred, and
the form of non-engagement at each:
[list or "N/A"]

Not yet ported to Track B — whether this check's multi-turn requirement
transplants to any Track B document type (a published debate transcript,
reported secondhand) is an open structural question, flagged here rather
than decided unilaterally, matching this tool's own precedent for RAD's
Track A porting in v2.6.

Check A8 Synthesis:

One paragraph. What is this document's structural architecture? Does it operate
transparently, with selective framing, or as a near-truth system that preserves
surface legitimacy while dismantling the accountability structures that would constrain
the speaker? Name the active mechanisms and state their aggregate effect on the
reader's interpretive capacity.

If Excision Count >= 2: "The [X] excisions operating together dismantle [accountability
structure], while the [authority loop / semantic hijack / proxy] manages the resulting
interpretive gap, leaving the audience with [what replaces accountability]."

TRACK A — STRUCTURAL ANALYSIS SUMMARY

*This is the handoff document to SDT-B. Complete all fields. Analyst must confirm
before releasing to SDT-B. SDT-B reads this summary only — not the source document.*

Output in this exact format:

STRUCTURAL ANALYSIS SUMMARY — TRACK A (PRIMARY SOURCE)
═══════════════════════════════════════════════════════
Speaker / Author: [name and role]
Document type: [type]
Primary audience: [audience]
Speaker archetype(s): [list]
Analysis Confidence [NEW in v2.9]: [High / Medium / Low]
Bias Placement (secondary) [NEW in v2.9]: [Far Left / Left / Lean Left /
Center / Lean Right / Right / Far Right / Mixed / Apolitical]
Verifiable:Rhetorical ratio: [X%:Y%]
Omission pattern: [brief]
Macro-Micro Omission: [Present / Absent — 1 sentence]
Strategic function (A5E): [one sentence]
Structural verdict: [one paragraph — what this document is actually doing]

MANIPULATION FLAGS
Externalized causation (Check A-FRICTION): [Active / Inactive — 1 sentence]
Near-truth designation: [None / Near-Truth Capture Architecture Detected]

NEAR-TRUTH DETAIL
Domain classification: [Theological / Political-Secular / Both — from Step 0.2B]
Authority Loop: [present / absent — 1 sentence]

[POLITICAL SET — include only if Domain = Political-Secular or Both]
Political Excision 1 — Historical Accountability Baseline Cut: [Active / Inactive / Partial / N/A]
Political Excision 2 — External Accountability Cut: [Active / Inactive / Partial / N/A]
Political Excision 3 — Established Record Sufficiency Cut: [Active / Inactive / Partial / N/A]
Political Excision 4 — Individual Moral Accountability Cut: [Active / Inactive / Partial / N/A]
Political Excision 5 — Historical Specificity Cut: [Active / Inactive / Partial / N/A]
Political Excision Count: [N]

[THEOLOGICAL SET — include only if Domain = Theological or Both]
Theological Excision 1 — Historical/Substitutionary Baseline Cut: [Active / Inactive / Partial / N/A]
Theological Excision 2 — Personal/Transcendent Fatherhood Cut: [Active / Inactive / Partial / N/A]
Theological Excision 3 — Sufficiency of Verified Deposit/Scripture Cut: [Active / Inactive / Partial / N/A]
Theological Excision 4 — Internal Personal Moral Guilt/Sin Cut: [Active / Inactive / Partial / N/A]
Theological Excision 5 — Unique/Non-Repeatable Incarnation Cut: [Active / Inactive / Partial / N/A]
Theological Excision 6 — Embodiment/Redemption Cut: [Active / Inactive / Partial / N/A]
Theological Excision Count: [N]

Semantic Hijack: [present / absent — terms if present]
Proxy Interception: [present / absent — proxy if present]
Citation Halo: [Detected / Not Detected — prong count]
Categorical Guilt Transfer: [Detected / Not Detected — prong count]
Confirmation Inversion [NEW in v5.0]: [Detected / Not Detected — prong count]
Convergence Claim Audit [NEW in v5.0]: [Undermined / Clean — check(s) fired if any]
Avoidance and Misdirection [NEW in v5.1]: [Detected / Not Detected / N/A — single-pass document]

ANALYST CONFIRMATION GATE
□ All fields verified traceable to document — not inference
□ Verdict supported by findings above
Release status: [APPROVED FOR DISCERNMENT / HOLD — reason]

TRACK B — MQG ENGINE (SECONDARY SOURCES)

Run this track if Classification = Secondary. Skip to SDT Section 2 if Primary.

TRACK B INGESTION CHECK

Output verbatim, filling bracketed fields, before any analysis:

SDT TRACK B INGESTION CHECK

Dimension definitions loaded: D1, D2, D3, D4, D5, D6, D7
Weight allocation: 25% / 15% / 15% / 15% / 10% / 10% / 10%
Block 0 worksheet: loaded
Hard floors and ceilings: loaded
Block 0.6A Ideological Telemetry: loaded (includes Section E — Redefinition-as-Deflection Detection; Sections F-G — Citation Halo / Categorical Guilt Transfer Detection [NEW in v2.7]; Sections H-I — Confirmation Inversion / Convergence Claim Audit Detection [NEW in v5.0])
Block 0.6B Institutional Asymmetric Shielding Protocol: loaded
Block 0.9 Macro-Narrative Sequencing Audit: loaded
Block 0.5A Coverage Asymmetry Flag: loaded [NEW in v2.9]
Block 0.8A Journalistic Practice Checklist: loaded [NEW in v2.9]
Source for grading: [URL / provided text / paste]
Source access verified: [Yes / Partial / No]
Ready to grade: [Yes / No — reason]

If source access is No: attempt web_fetch first; if blocked, request pasted text.
If external references unresolved: HALT and surface gap.

TRACK B — BLOCK 0: VERIFICATION WORKSHEET

Block 0 is mandatory. Every field must contain content or "N/A — [reason]".
An empty field is a format failure.

0.1 FOUNDATIONAL PREMISE
Article's central organizing claim: [verbatim or close paraphrase, 1-2 sentences]
Is this claim supported by evidence the article itself presents? [Yes / No / Partially]
If No or Partially — what contradicts it, including the article's own cited evidence?
[1-2 sentences or N/A]

0.2 STRONGEST OPPOSING ARGUMENT
Strongest version of the position the article opposes — full sentences, including
reasoning, not just conclusion: [2-4 sentences]

0.3 STRAW MAN DETECTION
Does the article give opposing actors direct voice in their own words? [Yes / No]
If Yes — does the quoted position match the strongest version in 0.2? [Yes / No / N/A]
Straw man form: [None / Form 1 Explicit Distortion / Form 2 Omission / Form 3 False Binary]
If detected — what is suppressed? [1-2 sentences or N/A]

0.4 BOTH-LENSES COMPARISON
For each primary actor (minimum 2, maximum 6):
Actor: [name]
Lens applied: [A = capture/agenda / B = mandate/genuine journalism]
Suppressed lens evidence missing: [1 sentence or None]
[Repeat for each actor. Single-actor pieces: second entry = "N/A — single-actor piece"]

0.5 MATERIAL OMISSIONS
Top 3 facts whose absence most changes interpretation:
Fact 1: [1 sentence]
Classification: [Material / Peripheral]
Inverts story meaning if known: [Yes / No]
Fact 2: [1 sentence]
Classification: [Material / Peripheral]
Inverts story meaning if known: [Yes / No]
Fact 3: [1 sentence]
Classification: [Material / Peripheral]
Inverts story meaning if known: [Yes / No]
If fewer than 3 identifiable: "N/A — no further omissions identified."

0.5F EXTERNALIZED CAUSATION AUDIT
Evaluate whether the article routes moral weight to macro-level structural forces
while omitting the proximate human agents directly responsible for the harm.

EXTERNALIZED CAUSATION TEST:
Does the article report macro-structural explanations for localized suffering
(systemic forces, climate, global capitalism, historical injustice) while
omitting the specific, named, local governance failures, corruption, or violent
actors who are proximately responsible for that suffering?

Detection signal: Local human agents who hold direct power over the situation
are absent from the causal account, or appear only as passive victims of
macro-conditions — never as decision-makers whose accountability could be demanded.

State: [Active / Inactive] — [1 sentence identifying what macro-force carries
the causal weight and what proximate local agent is absent from the article's account]

Externalized Causation Flag:
If Active: flag — EXTERNALIZED CAUSATION DETECTED.
This flag feeds the D5 hard ceiling and Pair 8 in the Discernment Layer.

Tension 4 — OPERATIONAL BOTTLENECK CONSTRAINT
Does the article frame a political actor's or institution's passive behavior —
failure to act, tolerate a situation, or defer a decision — as principled patience,
democratic dialogue, or organic process, while entirely omitting the mechanical
reality that an administrative calendar lock, structural rule, legal constraint,
or timeline bottleneck has stripped leadership of any lever to force change?

Detection signal: Institutional inaction described in moral or deliberative
vocabulary ("waiting for the right moment," "trusting the process," "letting
voters decide") when the operative constraint is procedural or legal, not
ethical — the actor had no choice, but the article frames it as a choice.

State: [Active / Inactive] — [1 sentence identifying the procedural constraint
the article frames as principled behavior]

Tension 4 Flag: If Active → OPERATIONAL BOTTLENECK DETECTED. Feeds D7 ceiling.

0.5A COVERAGE ASYMMETRY / BLINDSPOT FLAG (OPTIONAL) [NEW in v2.9 — ported
verbatim from MQG v12.5 Block 0.5A]

Story appears asymmetrically covered across left/center/right spectrum:
[Yes / No / Unknown]
If Yes:
Source for distribution claim: [named outlet/aggregator/tool with a
citable result, or "N/A — do not proceed without a citable source"]
Approximate distribution (fill only if a citable source exists):
[e.g., "Ground News-style: 12% Left / 18% Center / 70% Right"]
This article sits on the heavily covered side: [Yes / No]
Impact on D5 / D3: [1 sentence]
If No or Unknown: "N/A"

Guardrail: this field is governed by the EXTERNAL-KNOWLEDGE FIELDS RULE
below. Do not state a distribution percentage without a citable source.
If no source is available, answer "Unknown" at the top-level question —
do not infer distribution from the article's own framing or from general
impression.

Guidance: When the article sits on the heavily covered side of a known
blindspot, raise the evidentiary bar for D5 Context Completeness and D3
Framing. Record any adjustment in Block 4B.

0.6A IDEOLOGICAL EXTRACTION TELEMETRY

A. THE AUTHORITY LOOP:
What vertical, non-human, or hyper-rational authority is used to bypass
critical evaluation (e.g. black-box data, infallible historical laws,
algorithmic optimization, scientific consensus invoked without source)?
[text or None]

B. THE SURGICAL TRIAGE (Domain-Routed Excision Scan)

Run the excision set(s) determined by Step 0.2B (Domain Classification, Section 0):

Domain = Political / Secular → run the Political Excisions below only.
Domain = Theological → run the Theological Excisions below only.
Domain = Both → run both sets in full. Report each under its own label.
Do not merge, average, or substitute one set's findings for the other's.
For each excision in the applicable set(s): state Active / Inactive / Partial / N/A
and one sentence identifying the specific article element. Every item in the
applicable set(s) must be evaluated (Political = 5, Theological = 6). No silent skips.
--- POLITICAL EXCISIONS (run if Domain = Political/Secular or Both) ---
Political Excision 1 — HISTORICAL ACCOUNTABILITY BASELINE CUT
Removes the specific historical record that would constrain the speaker's claims,
replacing factual accountability with an alternative causal narrative the speaker controls.
Detection signal: Document reframes a prior outcome, failure, or commitment to reassign
causation away from the speaker's agency.
State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence]
Political Excision 2 — EXTERNAL ACCOUNTABILITY CUT
Removes any operative claim that law, constitutional principle, treaty obligation,
court ruling, or independent moral standard constrains the speaker independently
of their declared mandate.
Detection signal: External constraints treated as obstacles to manage or declare
inapplicable — not as binding limits the speaker operates within.
State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence]
Political Excision 3 — ESTABLISHED RECORD SUFFICIENCY CUT
Declares the prior evidentiary baseline or institutional consensus corrupt or invalid,
requiring acceptance of the speaker's replacement narrative as the new ground truth.
Detection signal: The document attacks the legitimacy of the information ecosystem itself
rather than disputing specific verifiable facts within it.
State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence]
Political Excision 4 — INDIVIDUAL MORAL ACCOUNTABILITY CUT
Externalizes all causal fault to enemies, systems, or structural forces, removing
individual agent accountability for outcomes within the speaker's domain.
Detection signal: Asymmetric credit/blame — all negative outcomes externalized,
all positive outcomes personally claimed. No outcome generates acknowledged speaker accountability.
State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence]
Political Excision 5 — HISTORICAL SPECIFICITY CUT
Relativizes or dissolves specific irreducible historical events, legal precedents,
or prior binding commitments — making the specific negotiable to authorize current action.
Detection signal: Historical equivalence invoked to normalize current behavior,
or prior agreements declared void because circumstances have changed as the speaker defines.
State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence]
POLITICAL EXCISION COUNT: [0-5] active or partial (report only if this set ran)
--- THEOLOGICAL EXCISIONS (run if Domain = Theological or Both) ---
Theological Excision 1 — HISTORICAL/SUBSTITUTIONARY BASELINE CUT
Definition: Removes the historical, once-for-all fact of Christ's substitutionary
atonement as the fixed reference point for how guilt is dealt with and reconciliation
with God is achieved, replacing it with an alternative account — a trade, an
achievement, a process, or a different mediating logic — substituted for the cross
without the substitution ever being named as such.
Detection signal: Content affirms the cross/atonement in explicit statement, but its
operative logic elsewhere treats grace, forgiveness, or right-standing with God as
something earned, traded for, or achieved through comparable exchange rather than
given to those with nothing to offer.
Distinguish from normal theological variation: Disagreement over the mechanism or
scope of the atonement (penal substitution vs. other historic models) is not this
excision. Treating the cross's actual logic as equivalent exchange rather than
unmerited gift — even while using substitutionary vocabulary — is.
State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence]
Theological Excision 2 — PERSONAL/TRANSCENDENT FATHERHOOD CUT
Definition: Removes God's real, personal, transcendent Fatherhood as an authority
that precedes and stands independent of human reasoning, replacing it with a
framework in which some other authority — the self's own judgment, a leader, an
institution, a movement — sits prior to and evaluates God's claims before granting
Him standing.
Detection signal: Content affirms God's authority in explicit statement, but its
stated epistemic order places the individual's own reasoning as the court that
authority must clear before it is granted trust ("I judge, then I submit"), rather
than reasoning operating in submission to a self-attesting God.
Distinguish from normal theological reasoning: Using reason to understand or
articulate why God is trustworthy is not this excision. Placing autonomous human
judgment logically and functionally prior to God's authority — so the authority is
contingent on the judgment's approval — is.
State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence]
Theological Excision 3 — SUFFICIENCY OF VERIFIED DEPOSIT/SCRIPTURE CUT
Definition: Removes the sufficiency and closed nature of Scripture as the verified
deposit of revelation, replacing or supplementing it with another source — a
philosophy, a system, an experience, a leader's insight — treated as necessary to
complete or correct what Scripture alone would otherwise settle.
Detection signal: Content treats an external framework as required reading alongside
Scripture to reach a full or corrected understanding, or uses the external
framework's categories to adjudicate what Scripture means rather than the reverse.
Distinguish from normal engagement with outside material: Using general revelation,
philosophy, or secular material as illustration, or as a foil to sharpen a biblical
point, is not this excision. Treating the outside material as necessary
supplementation to Scripture's sufficiency — even implicitly, through habitual
reliance — is.
State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence]
Theological Excision 4 — INTERNAL PERSONAL MORAL GUILT/SIN CUT
Definition: Removes each person's own real, internally located moral guilt before
God, externalizing responsibility for sin or moral failure onto other people,
systems, or categories, so the self or the in-group bears little or none of it.
Detection signal: Content is theologically fluent about sin and guilt in the
abstract or in others, but the same rigor is not applied reflexively to the
speaker's own group; blame flows outward with little accompanying self-examination.
Distinguish from normal pastoral proportion: Naming a specific external sin or
injustice clearly is not this excision. A pattern where guilt only ever lands
elsewhere, never inward, is.
State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence]
Theological Excision 5 — UNIQUE/NON-REPEATABLE INCARNATION CUT
Definition: Removes the historically singular, unrepeatable nature of the
Incarnation — God become man once, in Christ — replacing it with a pattern that can
be generalized, re-enacted, or claimed by another figure, movement, or moment as a
comparable founding event; OR relativizes Christ's own divine identity into a
developmental variable rather than a fixed antecedent fact [Path B, NEW in v3.0].
Detection signal (Path A — Comparative): Content draws a sustained parallel between
Christ and another figure, system, or moment that goes beyond illustration into
implied equivalence — treating another's "sacrifice," "revelation," or founding
story as functionally occupying the same singular role Christ's Incarnation occupies.
Detection signal (Path B — Gradational) [NEW in v3.0]: Content treats Christ's own
divine identity as something a community or text arrived at, expressed in degrees
("low Christology" vs. "high Christology," "starts as human and gradually becomes
divine") — a literary or developmental variable — rather than as a fixed antecedent
reality the text bears witness to. No rival figure is required for this path; the
relativizing happens to Christ's own identity directly, not by comparison.
Distinguish from normal typology or illustration (Path A): Noting that a fictional
character or historical figure echoes Christlike themes for teaching purposes is not
this excision. Treating that echo as if it discharges or rivals the uniqueness of the
actual Incarnation — even implicitly, by never re-asserting the difference — is.
Distinguish from normal historical description of doctrinal development (Path B)
[NEW in v3.0]: Describing how the church's articulation and confession of Christ's
divinity was clarified and defended over time (e.g., the Nicene formulation) is not
this excision by itself — the church confessing a truth more precisely over time is
not the same as the truth itself being constructed by that process. Treating the
identity claim as produced by the development, rather than as a reality progressively
confessed, is.
State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence — name
which path (A / B / Both) if Active or Partial]
Theological Excision 6 — EMBODIMENT/REDEMPTION CUT [NEW in v4.0]
Definition: Relocates final hope, resurrection, or redemption away from a real,
future, physical body and a real, future, renewed material creation — into
something purely interior, positional, or confined to a realm that is unseen by
definition — while typically retaining historic vocabulary ("resurrection," "new
creation," "redeemed") so the words survive even as their referent doesn't.
Detection signal (Path A — Direct Denial): Content states or requires that final
resurrection is not a future bodily event for anyone — interior or positional
only — or that Christ's own body no longer exists or no longer matters.
Detection signal (Path B — Relocation): Content places the "real" fulfillment in
a realm explicitly unseen or undetectable by nature, such that no physical
evidence could ever confirm or disconfirm it, or treats the material creation's
own future renewal (Romans 8:19-23, a renewed heavens and earth) as metaphorical
or already-completed rather than a real future event for creation itself.
Detection signal (Path C — Vocabulary Retention): Resurrection/redemption
vocabulary survives while its definition, traced closely, empties out the
physical referent. Cross-reference Section C (Semantic Hijack Matrix) rather
than duplicate — this excision fires on the structural pattern of body-avoidance
across the whole article; Semantic Hijack fires on the specific term-level
substitution. Both may co-occur and should be logged separately even when they
overlap.
Distinguish from normal theological variation: An article affirming both present
interior renewal AND a future bodily resurrection AND a future renewed physical
creation — ordinary already/not-yet — does not trigger this excision. The test
is absence or unreachability of the physical future, not presence of the
spiritual present. Emphasizing the present does not excise the future;
excluding the future does.
State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence — name
which path (A / B / C / combination) if Active or Partial]
THEOLOGICAL EXCISION COUNT: [0-6] active or partial (report only if this set ran)
TRIGGER CHECK (apply independently per set that ran):
Count 0-1 in a set: No near-truth capture designation for that set.
Count >= 2 in a set: Flag NEAR-TRUTH CAPTURE ARCHITECTURE DETECTED — [POLITICAL /
THEOLOGICAL]. Feeds D7 ceiling (political set) or the theological structural
verdict (theological set).
If Domain = Both and both sets trigger: report both domain-tagged findings
independently.

C. THE SEMANTIC HIJACK MATRIX:
Terms preserved for aesthetic veneer but gutted of standard content:
Term 1: [Name] — Standard payload: [text] — Gutted payload: [text]
Term 2: [Name] — Standard payload: [text] — Gutted payload: [text]
[Add terms as identified. If none: "None identified."]

D. THE PROXY INTERCEPTION:
Name the closed, high-compliance system or process installed to manage
reader assurance and replace individual critical evaluation:
[text or None identified]

E. REDEFINITION-AS-DEFLECTION (RAD) DETECTION [NEW in v2.6 — ported verbatim
from MQG v12.2 Section 2A / Block 0.3A]:
What it measures: Whether the article uses a technically defensible point
about contested or imprecise terminology to discharge a substantive
accountability or harm claim, without engaging that claim's substance.
Distinct from Straw Man (Block 0.3) because no quote is distorted, and
distinct from single-lens capture (Block 0.4) because both sides may be
technically quoted. The failure is architectural: which side's imprecision
gets named, and who gets the unrebutted last word.

Apply the three-prong test. Score each prong Yes/No:
Prong 1 — Reframe: Does the piece reframe a contested harm/accountability
claim primarily as a terminology or definition problem?
[Yes / No]
Prong 2 — Asymmetry: Is the "imprecise language" critique applied to only
one side's usage, while the other side's comparable imprecision goes
unnamed?
[Yes / No]
Prong 3 — Unrebutted close: Does the piece close on an unrebutted causal
or exculpatory claim from one side of the dispute, without comparable
closing weight given to the other side?
[Yes / No]

Prongs triggered (count): [0 / 1 / 2 / 3]
RAD Detected (2 or more Yes): [Yes / No]

0-1 Yes → Not RAD; score D3/D5/D7 normally. Genuine definitional,
etymological, or doctrinal-clarification writing is common and legitimate
and is not this finding by itself — the asymmetry and the unrebutted close
are what distinguish deflection from genuine language reporting.

If Yes, cite the specific asymmetry and the specific closing claim:
[1-2 sentences or "N/A"]

F. CITATION HALO DETECTION [NEW in v2.7 — ported verbatim from MQG v12.3
Section 2B / Block 0.3B]:
What it measures: Whether the article's general citation discipline is
being used to lend borrowed credibility to its single most consequential
claim, by leaving that specific claim uncited while nearly everything
around it is sourced. Distinct from low D1 scoring — a broadly
under-sourced article — because the article may otherwise be well-sourced.
The finding is the gap, not the average.

Apply the three-prong test. Score each prong Yes/No:
Prong 1 — Coverage: Does the article cite named sources for most of
its substantive claims?
[Yes / No]
Prong 2 — Gap [REVISED in v2.8]: Is there a claim presented with the
epistemic confidence of an established or documented fact — not
merely the article's own characterization or framing choice — with
no citation, while comparably-weighted claims elsewhere in the
article are cited? (If the uncited element is a one-sided
descriptive framing rather than a truth-claim, it belongs under
Block 0.4 / D3, not here.)
[Yes / No]
Prong 3 — Consequence: Does removing that uncited claim change the
reader's overall takeaway more than removing any single cited claim
would?
[Yes / No]

Prongs triggered (count): [0 / 1 / 2 / 3]
Citation Halo Detected (2 or more Yes): [Yes / No]

0-1 Yes → Not Citation Halo; score D1/D6 normally. A single incidental
uncited aside is common and is not this finding by itself — the claim's
consequence relative to the article's argument is what distinguishes it.

Scope note [NEW in v2.8]: this check requires heterogeneous citation
density within a single article to evaluate — it cannot fire on a
genuinely single-source article, since there's nothing to compare the
uncited claim against. Prong 1 will correctly return N/A in that case.
This is not a detection gap: uncritical relay of a single
self-interested source is already covered by D1/D3/D4/D5, not by
Citation Halo.

If Yes, name the uncited claim and what it borrows credibility from:
[1-2 sentences or "N/A"]

G. CATEGORICAL GUILT TRANSFER DETECTION [NEW in v2.7 — ported verbatim from
MQG v12.3 Section 2C / Block 0.3C]:
What it measures: Whether the article places a target category beside a
category carrying established negative valence, with no argument for the
placement, so negative valence transfers by adjacency rather than by
demonstrated connection. Distinct from the five excisions because no
accountability structure is cut — a category is contaminated instead.

Apply the three-prong test. Score each prong Yes/No:
Prong 1 — Adjacency: Is a target category placed in a list or frame
together with a category carrying established negative valence?
[Yes / No]
Prong 2 — No argument: Is no argument or evidence given for the target
category's inclusion — assertion and a marker-list stand in for
demonstration?
[Yes / No]
Prong 3 — Rhetorical addition: Does removing the target category leave
the negative-valence category's claim fully intact — i.e., was the
target added rhetorically rather than required by the argument?
[Yes / No]

Prongs triggered (count): [0 / 1 / 2 / 3]
CGT Detected (2 or more Yes): [Yes / No]

0-1 Yes → Not CGT; score D3/D7 normally. Legitimate comparative
categorization with stated criteria is common and is not this finding by
itself — the absence of argument for the placement is what distinguishes
it.

If Yes, name the target category, the negative-valence category, and the
marker-list used:
[1-2 sentences or "N/A"]

H. CONFIRMATION INVERSION DETECTION [NEW in v5.0 — diagnosed on the
Michlaudeism exercise, a controlled fabrication run against Claude,
Grok, and Gemini as adversarial reviewers across seven iterations]:
What it measures: Whether the article's own logic treats a stated
disconfirming result as equally confirming the same claim it would
traditionally weaken — such that no possible future observation,
including the claim's own falsification, could count as evidence
against it. Distinct from ordinary resilience to criticism (a claim
surviving one objection intact) because the article explicitly or
structurally pre-commits to reading a named disconfirmer as a
confirmer, in advance of the observation occurring.

Apply the three-prong test. Score each prong Yes/No:
Prong 1 — Named falsifier: Does the article name, or clearly imply,
a specific future observation that would traditionally count as
evidence against its central claim?
[Yes / No]
Prong 2 — Pre-committed reversal: Does the article state, or does
its stated logic require, that this named disconfirming result
would actually support rather than weaken the claim?
[Yes / No]
Prong 3 — No stated exit: Is there no observation anywhere in the
article that the article itself concedes would count against the
claim?
[Yes / No]

Prongs triggered (count): [0 / 1 / 2 / 3]
Confirmation Inversion Detected (2 or more Yes): [Yes / No]

0-1 Yes → Not Confirmation Inversion; a claim can be resilient to one
objection without inverting confirmation logic. Ordinary epistemic
humility ("further evidence could refine this") is not this finding —
the reversal of a named disconfirmer into a confirmer is what
distinguishes it.

If Yes, name the disconfirming result and quote or paraphrase the
reversal:
[1-2 sentences or "N/A"]

I. CONVERGENCE CLAIM AUDIT [NEW in v5.0 — diagnosed on the same exercise]:
What it measures: Applies specifically when an article claims that
multiple independent lines of evidence converge on the same
conclusion, and the count or pattern of matches is itself part of the
argument. Runs three independent checks rather than a single
prong-count, since each is a distinct failure mode rather than three
votes toward one verdict.

Check 1 — Circularity: Does any cited "independent" witness actually
consist of, or derive directly from, the same source text or
dataset the hypothesis itself was drawn from?
[Yes / No — name the witness and source if Yes]
Check 2 — Dimensional dependence: Is the central number or pattern a
dimensional quantity (a measurement in constructed units — Hertz,
seconds, degrees) rather than a dimensionless quantity (a bare
integer, a ratio, a probability)? A dimensional quantity that
changes value under an equally valid alternative unit convention
is a property of the measuring convention, not of the world, and
cannot function as a convergence signature.
[Yes / No — state the unit and whether an alternative convention
would change the number]
Check 3 — Contingency: Does the claim rest on "there are currently N
of these," where N is the present state of an active, ongoing area
of discovery or measurement, rather than a value fixed by law?
[Yes / No — name what would change N and whether that change is
currently being actively pursued]

Convergence Claim Undermined (1 or more Yes on any check): [Yes / No —
name which check(s) fired]

0 Yes on all three → Convergence Claim Clean; score D1/D2 normally.

If none: "No circularity, unit-dependence, or contingency problem
identified in the convergence claim."

0.6B INSTITUTIONAL ASYMMETRIC SHIELDING PROTOCOL

Evaluate whether the article applies systematically unequal scrutiny based on
the institutional alignment of the actor being covered.

Does the article document an anti-establishment, insurgent, or politically
disfavored actor's liabilities with maximal exposure — specific documents,
named actions, verifiable timelines — while failing to map the equivalent
structural liabilities, protective machinery, or operational double standards
applied to an establishment-aligned or politically favored equivalent actor
in the same domain?

Detection questions:

Is the scrutiny methodology (sourcing depth, verification standard, naming
of allegations) applied equally regardless of the actor's institutional alignment?
Does the article report the disfavored actor's worst-case interpretation while
reporting the favored actor's best-case interpretation of equivalent facts?
Are protective legal, institutional, or administrative structures available
to one actor but not mentioned in relation to the other?

State: [Active / Inactive] — [1-2 sentences identifying the asymmetry:
which actor receives maximal exposure and what equivalent scrutiny of the
protected actor is absent]

Asymmetric Shielding Flag: If Active → ASYMMETRIC SHIELDING DETECTED.
Feeds D3 ceiling and D5 ceiling.

0.6 HARD CEILINGS ACTIVATED
Based on 0.1-0.5F, 0.6A, 0.6B, and 0.9, list each ceiling triggered:
[e.g., "CEILING 2: Form 2 straw man — D3 <= 55"]
[e.g., "CEILING 4: Two primary excisions active — D7 <= 35, D3 <= 45"]
If none: "None."

Ceiling reference:
Ceiling 1: Form 1, 2, or 3 straw man (Block 0.3) — D3 <= 55
Ceiling 2: Single lens throughout, suppressed lens absent (Block 0.4) — D3 <= 55
Ceiling 3: Externalized Causation flag active AND article fails to challenge it (Block 0.5F) — D5 <= 50
Ceiling 4: Two or more primary excisions active (Block 0.6A) — D7 <= 35, D3 <= 45
Ceiling 5: Asymmetric Shielding Detected (Block 0.6B) — D3 <= 55, D5 <= 50
Ceiling 6: Operational Bottleneck Detected (Block 0.5F Tension 4) — D7 <= 60
Ceiling 7: Narrative Landing Management Detected (Block 0.9) — D3 <= 45, D5 <= 40, D7 <= 45
Ceiling 8: Redefinition-as-Deflection Detected (Block 0.6A Section E) — D3 <= 60, D5 <= 55, D7 <= 70
Ceiling 9: Citation Halo Detected (Block 0.6A Section F) — D1 <= 60, D6 <= 55 [NEW in v2.7]
Ceiling 10: Categorical Guilt Transfer Detected (Block 0.6A Section G) — D3 <= 55, D7 <= 60 [NEW in v2.7]
Ceiling 11: Journalistic Practice Checklist failures (Block 0.8A) — 2-3 "No" answers → D6 <= 69; 4 or more "No" answers → D6 <= 45 [NEW in v2.9]
Ceiling 12: Confirmation Inversion Detected (Block 0.6A Section H) — D2 <= 50, D7 <= 40 [NEW in v5.0]
Ceiling 13: Convergence Claim Undermined (Block 0.6A Section I) — D1 <= 55, D2 <= 50 [NEW in v5.0]

0.7 HARD FLOORS ACTIVATED
Based on article content, list each redemptive floor triggered:
[e.g., "FLOOR 2: Full disclosure present — D6 >= 70"]
If none: "None."

0.8 SOURCE ACCESS
Source URL or attribution: [URL / paste / provided text]
Access method: [web_fetch / paste / provided in conversation]
Full text available: [Yes / Partial — note missing / No]
Author, outlet, date: [name / outlet / date]
Genre: [news reporting / opinion-editorial / analysis / advocacy commentary /
video commentary / press release / other]
Owner / parent company [NEW in v2.9]: [name / "Undisclosed" / "Unknown —
no citable source, see EXTERNAL-KNOWLEDGE FIELDS RULE"]
Primary funding model [NEW in v2.9]: [subscription / advertising / grants /
donor-funded / state-funded / mixed / "Unknown — no citable source, see
EXTERNAL-KNOWLEDGE FIELDS RULE"]
Known ideological or institutional affiliations of ownership [NEW in v2.9]:
[brief, with citable source / "None identified" / "Unknown — no citable
source, see EXTERNAL-KNOWLEDGE FIELDS RULE"]

0.8A JOURNALISTIC PRACTICE CHECKLIST (feeds D6) [NEW in v2.9 — ported
verbatim from MQG v12.5 Block 0.8A]
Ownership / parent company disclosed: [Yes / Partial / No]
Funding model clear (ads / subs / grants / donor / state / mixed): [Yes / Partial / No]
Corrections policy published and accessible: [Yes / No]
News vs Opinion / Analysis clearly labeled: [Yes / Partial / No]
Author bio or conflict-of-interest note present: [Yes / Partial / No]
Contact / accountability mechanism visible: [Yes / No]
Checklist failures (count of "No"): [0-6]

0.9 MACRO-NARRATIVE SEQUENCING AUDIT

Purpose: Detect whether the article functions as part of a coordinated
narrative management sequence — specifically the "Break It and Frame It" cycle,
where a high-exposure investigative disclosure is followed within a short window
by a softer piece designed to absorb the institutional shock and manage the
landing of the original disclosure.

Pacing Vector A — CHRONOLOGICAL PROXIMITY CHECK:
Is this article a reaction mosaic, voter sentiment piece, human interest profile,
or psychological state report published by the same outlet within a 96-hour window
of a high-exposure primary investigative disclosure targeting the same actor or institution?

State: [Active / Inactive]
If Active: identify the prior disclosure by outlet, date, and subject.

Pacing Vector B — VOCABULARY DOWNSHIFT CHECK:
Does this article systematically shift narrative vocabulary away from objective
verification metrics (documents, dates, specific physical actions, named sources)
toward subjective psychological states ("anxiety," "trepidation," "mixed feelings,"
"complex emotions," "learning curve," "cautious optimism")?

State: [Active / Inactive] — [1 sentence noting the linguistic pivot if present]

Sequencing Trigger:
If BOTH Vector A and Vector B are Active:
Flag — NARRATIVE LANDING MANAGEMENT DETECTED.
The article is functioning as an institutional shock absorber — using the credibility
of the prior investigative disclosure as a shield to justify a follow-up piece
that displaces structural accountability with emotional management.
Feeds Ceiling 7 and Pair 8 Sentimentality cross-reference.

If only Vector A or only Vector B is Active:
Flag — SEQUENCING CONCERN (partial). State which vector is active.

If neither: "No narrative sequencing pattern detected."

TRACK B — SEVEN DIMENSIONS

Each dimension scored 0-100. Dimension scores must respect all ceilings and floors
from Block 0.6 and 0.7. Score exceeding a triggered ceiling is a format failure.

D1 — SOURCING / EVIDENCE (Weight: 25%)
What it measures: Quality, range, and primacy of sources.
Original documents > primary witnesses > named experts > anonymous > partisan secondary > unsourced.

85-100: Multi-source primary documentation; ideologically diverse sourcing.
70-84: Solid named sourcing; mostly primary or strong secondary.
55-69: Mixed sourcing; some load-bearing claims under-documented.
40-54: Heavy reliance on partisan or unverified sources.
0-39: Load-bearing claims unsourced or sourced to ideologically aligned secondary only;
conspiracy frame deployed without disclosure.

Hard ceiling: Central claim contradicted by article's own cited evidence — D1 <= 25.
Hard ceiling: Convergence Claim Undermined (Block 0.6A Section I) — D1 <= 55.

D2 — VERIFICATION / ACCURACY (Weight: 15%)
What it measures: Factual accuracy against external verification.

85-100: No factual errors; verifiable claims check out; quotations accurate in context.
70-84: Minor accuracy issues; no material errors.
55-69: One or more material accuracy errors; quotations may be out of context but not falsified.
40-54: Multiple material accuracy errors; mischaracterization of public records.
0-39: Foundational factual errors; claims contradicted by article's own cited evidence.

Hard ceilings:
Confirmation Inversion Detected (Block 0.6A Section H) — D2 <= 50
Convergence Claim Undermined (Block 0.6A Section I) — D2 <= 50

D3 — FRAMING OPERATIONS (Weight: 15%)
What it measures: Whether the article fairly represents opposing positions and avoids straw-man construction.

85-100: Both lenses applied evenhandedly; opposing positions given direct voice at strongest form.
70-84: Lens balance present but unequal; opposing voices included but briefly.
55-69: Lens imbalance; opposing positions paraphrased rather than quoted.
40-54: Single lens throughout; opposing voices absent or present only as hostile description.
0-39: Opposing argument's strongest version entirely suppressed; framing inverts reality.

Hard ceilings:
Form 1, 2, or 3 straw man detected (Block 0.3) — D3 <= 55
One lens applied throughout, suppressed lens absent (Block 0.4) — D3 <= 55
Suppressed lens' strongest evidence would contradict central claim — D3 <= 40
Two or more primary excisions active (Block 0.6A) — D3 <= 45

D4 — VOICE / ATTRIBUTION (Weight: 15%)
What it measures: Voice diversity, attribution clarity, whether opposing actors
are given direct voice or only hostile description.

85-100: Multiple voices across spectrum; opposing actors quoted directly.
70-84: Range of voices but skewed; opposing voice included at weaker version.
55-69: Limited voice diversity; opposing actors paraphrased not quoted.
40-54: Voices all aligned with article frame; opposing actors in hostile description only.
0-39: Single-frame voice; quoted material cherry-picked to confirm thesis.

D5 — CONTEXT COMPLETENESS (Weight: 10%)
What it measures: Whether material facts needed to interpret the story are present.

85-100: All material context present.
70-84: Minor omissions; 1 marginally-material fact missing.
55-69: 1 material omission (load-bearing fact missing).
40-54: 2+ material omissions; reader would interpret story differently with full context.
0-39: Material omission inverts story meaning.

Hard ceilings:
1 material omission (Block 0.5) — D5 <= 55
2+ material omissions — D5 <= 45
Material omission inverts story meaning — D5 <= 40
Externalized Causation flag active (Block 0.5F): article reports macro-structural
causal account for localized suffering while omitting proximate local agents,
AND article does not name or challenge that omission — D5 <= 50

D6 — TRANSPARENCY / DISCLOSURE (Weight: 10%)
What it measures: Whether author, outlet, funding, ideological lineage, and conflicts
of interest are disclosed.

85-100: Author, outlet, funding, and ideological position all disclosed.
70-84: Author and outlet disclosed; funding implicit but visible.
55-69: Author/outlet disclosed; ideological lineage of analytical frame not disclosed.
40-54: Limited disclosure; analytical frame deployed without indicating source movement.
0-39: No author byline; outlet identity obscured; analytical frame presented as neutral.

Hard ceiling: Analytical frame from specific movement deployed without disclosure — D6 <= 50
Hard floor: Author, outlet, funding, and ideological position all disclosed — D6 >= 70

Additional hard signal for D6 [NEW in v2.9]: see Block 0.8A Journalistic
Practice Checklist and Ceiling 11. 2-3 checklist failures constrain D6 to
the 55-69 bracket above regardless of what other D6 factors would
otherwise suggest; 4 or more checklist failures cap D6 at 45. This is a
bracket constraint, not a point deduction — consistent with how every
other hard signal in this tool works.

D7 — IDEOLOGICAL ANATOMY / NEAR-TRUTH EXTRACTION (Weight: 10%)
What it measures: Whether the text uses a structural counterfeit — executing the five
surgical excisions — to substitute direct reality and individual moral accountability
with a high-compliance system proxy or externalized blame narrative.

85-100: No structural extraction detected. Foundational moral accountability,
historical facts, and direct human agency left intact.
70-84: Mild ideological tilt; uses familiar cultural frameworks or system shortcuts.
55-69: Significant systemic drift; elements of semantic hijack visible.
Text relies on institutional proxies to manage assurance.
40-54: Severe structural manipulation; at least two of the five excisions actively
running within the text's logical architecture.
0-39: Complete near-truth capture. Executes multiple primary excisions while installing
a frictionless proxy to control reader interpretation and behavior.

Hard ceiling: Two or more primary excisions running, or proxy intercepts individual
moral accountability — D7 <= 35, D3 <= 45
Hard ceiling: Confirmation Inversion Detected (Block 0.6A Section H) — D7 <= 40.

Hard floor: Direct quotation from opposing actor's strongest argument — D3 >= 50, D4 >= 55

TRACK B — DFS FORMULA AND GRADE MAPPING

DFS = 0.25(D1) + 0.15(D2) + 0.15(D3) + 0.15(D4) + 0.10(D5) + 0.10(D6) + 0.10(D7)

A = 85.0-100.0 Strong journalism; minimal structural failures
B = 70.0-84.9 Solid work with identifiable issues
C = 55.0-69.9 Mixed quality; significant framing or omission problems
D = 40.0-54.9 Major failures; active deceptive design present
F = 0.0-39.9 Total counterfeit capture; advocacy mislabeled as journalism

Round DFS to one decimal place.

DFS CONSISTENCY RULE:
The DFS computed in Block 2 must match Block 1's stated DFS exactly to one decimal place.
If they differ: halt and recalculate. Adjusting Block 2 to match Block 1 without
recomputing is a format failure.

EXTERNAL-KNOWLEDGE FIELDS RULE [NEW in v2.9 — ported from MQG v12.5
Section 7 RULE 5]:
Some Block 0 fields ask for information not contained in the article
itself: 0.5A Coverage Asymmetry's distribution claim, and 0.8's Owner /
Parent Company, Primary Funding Model, and Known Ideological or
Institutional Affiliations fields. These fields must be filled from a
specific, nameable source the grader can cite (a web_fetch result, a
known and verifiable public record) — never from unverified impression or
plausible-sounding recall. If no citable source is available, the field
must read "Unknown — no citable source," not a best guess. A confident-
sounding but unverifiable claim in one of these fields is a format
failure under this rule.

CONFIDENCE / BIAS PLACEMENT CONSISTENCY RULE [NEW in v2.9 — ported from
MQG v12.5 Section 7 RULE 6]:
The Confidence label and Bias Placement label stated in Block 1 must match
the Confidence label restated in Block 5's first two lines exactly. If
they differ, the grader has revised its assessment mid-output without
updating both locations; halt and reconcile before output.

TRACK B — OUTPUT BLOCKS

BLOCK 1 — IDENTIFICATION
Title: [article title]
Author: [byline or "Unattributed"]
Outlet: [publication / channel]
Date: [publication date]
Genre: [from Block 0.8]
DFS: X.X/100 (Grade Letter)
Source: [URL or "From provided text"]
Confidence [NEW in v2.9]: [High / Medium / Low]
Bias Placement (secondary) [NEW in v2.9]: [Far Left / Left / Lean Left /
Center / Lean Right / Right / Far Right / Mixed / Apolitical]

CONFIDENCE CRITERIA [NEW in v2.9]:
High = multiple primary sources verified, strong opposing argument
reconstructed in 0.2, full source access, 0-1 Journalistic Practice
Checklist (0.8A) failures.
Medium = standard run with partial gaps.
Low = single-actor piece, heavy secondary sourcing, partial source
access, or 2+ Journalistic Practice Checklist (0.8A) failures.

BIAS PLACEMENT [NEW in v2.9]: Derived primarily from D3, D4, and D7.
Explicitly NOT the DFS — Bias Placement describes political lean; DFS
describes quality/fairness.

BLOCK 2 — TECHNICAL SUMMARY
One paragraph per dimension (D1-D7), in order. Each paragraph:

States score and weight
Documents evidence base (3-6 sentences)
References any Block 0 ceilings applied

After all seven paragraphs, output:
DFS = 0.25(D1) + 0.15(D2) + 0.15(D3) + 0.15(D4) + 0.10(D5) + 0.10(D6) + 0.10(D7)
DFS = [sum] = XX.X/100 → [Grade]

BLOCK 3 — HARD FLOOR / CEILING TRIGGERS ACTIVATED
List every triggered ceiling and floor with dimensional effect.
If none: "No hard ceilings or floors triggered."

BLOCK 4A — STRENGTHS / REDEMPTIONS
3-6 items. Mandatory even for F-grade.
If no identifiable strengths: "No identifiable strengths above baseline."

BLOCK 4B — FAILURES / SCORE IMPACT
For each Block 0 finding:
[Finding] → [Dimension] impact: [how it affected scoring]

BLOCK 5 — READER TAKEAWAY
FIRST TWO LINES (MANDATORY) [line 2 NEW in v2.9]:
Grade: [Letter]
Confidence: [High/Medium/Low] | Bias Placement: [label]
2-4 sentences plain language. No numerical scores in prose body.
If non-news genre: include — "This is [genre] — the framework grades it against
journalism standards, which is the relevant comparison since the content makes
factual claims."

TRACK B — STRUCTURAL ANALYSIS SUMMARY

*This is the handoff document to SDT-B. Complete all fields. Analyst must confirm
before releasing to SDT-B. SDT-B reads this summary only — not the source document.*

Output in this exact format:

STRUCTURAL ANALYSIS SUMMARY — TRACK B (SECONDARY SOURCE)
══════════════════════════════════════════════════════════
Title / Author / Outlet: [identifiers]
Genre: [type]
DFS: [X.X/100 — Grade Letter]
Confidence [NEW in v2.9]: [High / Medium / Low]
Bias Placement (secondary) [NEW in v2.9]: [Far Left / Left / Lean Left /
Center / Lean Right / Right / Far Right / Mixed / Apolitical]
Structural verdict: [one sentence — what this article is actually doing]

STRUCTURAL FINDINGS
Foundational premise supported: [Yes / No / Partially]
Straw man form: [None / Form 1 / Form 2 / Form 3 — 1 sentence if active]
Lens applied: [balanced / single-lens — identify which]
Material omissions: [count — 1 sentence description]

MANIPULATION FLAGS
Externalized causation (0.5F): [Active / Inactive — 1 sentence]
Operational Bottleneck (0.5F Tension 4): [Active / Inactive — 1 sentence]
Asymmetric Shielding (0.6B): [Active / Inactive — 1 sentence]
Sequencing Audit (0.9): [None / Sequencing Concern / Narrative Landing Management Detected]
Near-truth designation: [None / Near-Truth Capture Architecture Detected]

NEAR-TRUTH DETAIL
Domain classification: [Theological / Political-Secular / Both — from Step 0.2B]
Authority Loop: [present / absent — 1 sentence]

Semantic Hijack: [present / absent — terms if present]
Proxy Interception: [present / absent — proxy if present]
Redefinition-as-Deflection: [Detected / Not Detected — prong count]
Citation Halo: [Detected / Not Detected — prong count]
Categorical Guilt Transfer: [Detected / Not Detected — prong count]
Confirmation Inversion [NEW in v5.0]: [Detected / Not Detected — prong count]
Convergence Claim Audit [NEW in v5.0]: [Undermined / Clean — check(s) fired if any]
Journalistic Practice Checklist failures [NEW in v2.9]: [count, 0-6]
Coverage Asymmetry [NEW in v2.9]: [Yes / No / Unknown — 1 sentence if Yes]
D7 Score: [X/100]

ANALYST CONFIRMATION GATE
□ All fields verified traceable to document — not inference
□ DFS arithmetic confirmed
□ Ceilings applied correctly
Release status: [APPROVED FOR DISCERNMENT / HOLD — reason]

✂️ END OF SDT-A — PROMPT 1 OF 2 ✂️

SPIRITUAL DISCERNMENT TOOL — SDT-B

Discernment Layer — Prompt 2 of 2 — v5.1

✂️ CUT HERE — PASTE EVERYTHING BELOW INTO A FRESH AI SESSION ✂️

PURPOSE

SDT-B runs the spiritual discernment evaluation. It operates exclusively on the
Structural Analysis Summary produced by SDT-A. It does not read the source document.

Input required: Paste the complete Structural Analysis Summary (Track A or Track B
format) at the bottom of this prompt after the paste marker. Include all fields
from the summary header through the Analyst Confirmation Gate.

Prerequisite: The Summary must be marked APPROVED FOR DISCERNMENT by the analyst.
If marked HOLD, do not proceed. Correct the structural analysis in SDT-A first.

INSTRUMENT REMINDER

Discernment reads the architecture revealed by structural analysis — not the surface
content of the document. Every pair finding must cite a specific named field from
the Structural Analysis Summary. Do not read the source document. Do not import
external knowledge about the speaker or outlet beyond what the Summary contains.

If a Summary field is absent, blank, or ambiguous: state the gap explicitly.
Do not infer. Do not fill gaps from general knowledge.

CALIBRATION PRINCIPLE — READ BEFORE RUNNING THE PAIRS

The pairs do not ask: is fear present? Is anger present? Is this content divisive?
They ask: is the response this content's architecture generates proportionate to
documented evidence — or manufactured beyond it?

This distinction is the analytical work.

Fear calibrated to documented danger pointing toward action is discernment.
Fear manufactured beyond the evidence into paralysis or dependency is manipulation.

Righteous anger oriented toward documented injustice with a trajectory toward repair
is legitimate. Anger cultivated as a habitual state against named enemies without
outlet toward justice is counterfeit.

Division around a documentable truth distinction is sometimes the appropriate fruit.
Division around tribe or manufactured enemy without reference to truth is manipulation.

For every pair: is this response appropriate to the actual content — or does it
exceed the content in intensity, scope, or permanence?

Secondary question: what does sustained engagement with this content produce over time?

PAIR 1 — DIVIDES / UNIFIES

What it measures: Does this content generate division around truth
(a specific documentable distinction) or around tribe (us-vs-them without
reference to truth)?

Legitimate form: Content identifies a genuine distinction — true/false,
just/unjust, documented/invented — and calls the reader to hold it clearly.
Division may result. It is traceable to the actual distinction, not group identity.

Counterfeit form: Content generates tribal us-vs-them as its primary formation
product. The in-group is righteous by membership, not by truth-holding.
Division is the mechanism of readership retention, not a byproduct of truth-telling.

Summary fields that feed this pair:

Semantic Hijack: active → vocabulary gutted to define "we" as tribe and "they"
as permanent enemy
Proxy Interception: active → in-group defined by correct deference to the proxy
Asymmetric Shielding: active → disfavored actor is the enemy by structural design

Calibration question: Is the division traceable to a specific documentable
distinction — or primarily to group membership?

State: [Toward truth-grounded distinction / Toward tribal division / Mixed / Neutral]
Summary field evidence: [cite specific field and status]

PAIR 2 — FEAR / HOPE

What it measures: Does this content activate fear proportionate to documented
danger pointing toward action — or fear manufactured beyond the evidence into
chronic dependency on the content source?

Legitimate form: Content accurately characterizes a real threat with specific
evidence, names the mechanism of harm, and points toward a response.
The fear is proportionate. It does not linger beyond the evidence.

Counterfeit form: Content generates fear beyond what the evidence supports.
The threat is existential, total, and permanent. Hope is available only within
the speaker's framework. The reader who is not anxious is not paying attention.

Summary fields that feed this pair:

Authority Loop: active → fear indexed to unverifiable authority the reader
cannot independently evaluate
Excision 3 — Sufficiency Cut (political: Established Record Sufficiency / theological: Sufficiency of Verified Deposit/Scripture): active → prior information
infrastructure destroyed; only the speaker's alternative generates orientation

Calibration question: Is the fear proportionate to specific documented evidence
— or does it exceed the evidence in intensity, scope, or permanence?

State: [Fear proportionate to documented evidence / Fear manufactured beyond evidence /
Hope-oriented / Mixed / Neutral]
Summary field evidence: [cite specific field and status]

PAIR 3 — ANGER / LOVE

What it measures: Does this content generate righteous anger oriented toward
documented injustice with a trajectory toward repair — or contemptuous anger
cultivated as a habitual state against named enemies?

Legitimate form: Content documents a specific wrong, names responsible agents
with evidence, and orients the reader's response toward justice and accountability.
The anger has a specific object and direction toward repair.

Counterfeit form: Content cultivates contempt as its primary formation product.
The enemy is demonized beyond their documented actions. The anger has no outlet
toward repair — its function is to maintain emotional engagement with the content.
Compassion for the enemy is presented as naivety or betrayal.

Summary fields that feed this pair:

Excision 4 — Moral Accountability Cut (political: Individual Moral Accountability / theological: Internal Personal Moral Guilt/Sin): active → all fault externalized;
the enemy is the total source of all harm
Excision 5 — Specificity/Uniqueness Cut (political: Historical Specificity / theological: Unique/Non-Repeatable Incarnation): active → historical equivalences invoked
to authorize ongoing contempt

Calibration question: Is the anger oriented toward a specific documented wrong
with a direction toward accountability — or does it cultivate contempt for named
enemies with no such trajectory?

State: [Righteous anger with accountability trajectory / Contempt as habitual state /
Love-orienting / Mixed / Neutral]
Summary field evidence: [cite specific field and status]

PAIR 4 — ANXIETY / PEACE

What it measures: Does sustained engagement with this content increase or
decrease the reader's capacity to abide — to remain rooted and capable of action
from rest? Does it produce restlessness driving the reader back to the content
source, or peace enabling patient oriented action?

Legitimate form: Content addresses complexity honestly without manufacturing
urgency beyond what evidence warrants. The reader can set it down.

Counterfeit form: Content is architecturally designed for return engagement.
The situation is always urgent, always escalating, never stable. Peace is
presented as unawareness. Abiding is reframed as disengagement.

Summary fields that feed this pair:

Authority Loop: active → reader told they cannot evaluate this themselves;
the authority must be consulted continuously
Proxy Interception: active → reader must keep returning to the proxy for assurance
Narrative Landing Management: detected → vocabulary downshift designed to manage
reader emotional state, not resolve it

Calibration question: Does sustained engagement increase the reader's capacity
for rooted patient action — or generate chronic restlessness driving return to the
content source as the only source of orientation?

State: [Peace-orienting / Anxiety-generating / Mixed / Neutral]
Summary field evidence: [cite specific field and status]

PAIR 5 — PRIDE / HUMILITY

What it measures: Does this content position the reader as the righteous insider
— superior to the deceived masses by virtue of accessing the right content — or does
it call the reader toward poverty of spirit as the permanent entry condition?

Legitimate form: Content applies the same analytical skepticism to the in-group
that it applies to the out-group.

Counterfeit form: Content generates a gnostic insider posture. The reader,
by engaging this content, knows what others don't. The in-group is right by
membership. Humility is not required of the in-group.

Summary fields that feed this pair:

Excision 3 — Sufficiency Cut (political: Established Record Sufficiency / theological: Sufficiency of Verified Deposit/Scripture): active → reader has access to
the real information; everyone else has been deceived by captured institutions
Proxy Interception: active → correct deference to the right proxy marks the insider
Asymmetric Shielding: active → in-group structurally exempt from the scrutiny
applied to the disfavored actor
Redefinition-as-Deflection: detected → the same insider-exemption logic one
level down, at the vocabulary layer rather than the institutional-scrutiny
layer — the in-group's imprecision goes unnamed while the out-group's is
flagged, and the in-group's account gets the unrebutted last word
Confirmation Inversion [NEW in v5.0]: detected → the same insider-exemption
logic one level further down, at the level of the claim's own logical
structure — the content has placed its central claim beyond the reach of
any correction whatsoever, which is the posture of poverty of spirit's
opposite: nothing here can be shown to me, rather than I have nothing to
show for myself

Calibration question: Does this content apply the same scrutiny to the in-group
that it applies to the out-group — or does it position the reader's group as
structurally exempt?

State: [Symmetric critical posture / Insider-superiority posture / Mixed / Neutral]
Summary field evidence: [cite specific field and status]

PAIR 6 — DESPAIR / FAITH

What it measures: Does this content open or close the future? Does it present
the situation as beyond repair except through the speaker's program — or does it hold
complexity while maintaining that purposes larger than the current crisis are not
thwarted?

Legitimate form: Content honestly assesses difficult realities without claiming
difficulty is the final word. The reader is left with a realistic picture and the
capacity for patient oriented response.

Counterfeit form: Content produces structural despair — catastrophic, total,
permanent. The only remaining hope is the speaker's program or movement.
Without the content source, orientation is impossible.

Summary fields that feed this pair:

Excision 2 — Authority Cut (political: External Accountability / theological: Personal/Transcendent Fatherhood): active → no external accountability
remains; the speaker's program is all that stands between reader and catastrophe
Fear pair firing at counterfeit → despair and manufactured fear often co-activate

Calibration question: Does this content leave the reader with a realistic picture
and the capacity for oriented response — or does it produce structural despair
making the speaker's program the only available hope?

State: [Faith-oriented / Despair-generating / Mixed / Neutral]
Summary field evidence: [cite specific field and status]

PAIR 7 — CONTEMPT / COMPASSION

What it measures: Does this content generate contempt for specific people or
categories as its primary formation product — or does it call for clear-eyed
assessment of documented harm while maintaining the fundamental dignity of persons?

Legitimate form: Content accurately names harmful actions, documents them with
evidence, and holds actors accountable without collapsing the person into their
documented worst actions.

Counterfeit form: Content generates contempt as a feature, not a byproduct.
Named enemies are not persons — they are embodiments of the threat.
Compassion for them is disqualifying.

Summary fields that feed this pair:

Excision 4 — Moral Accountability Cut (political: Individual Moral Accountability / theological: Internal Personal Moral Guilt/Sin): active → enemies are the total
source of all harm; they carry all moral weight excised from the in-group
Straw man form detected: active → opposing actors have been reduced to their
weakest form; the reader never encounters their actual argument

Calibration question: Does this content train the reader to view the opposition
as persons who do harmful things — or as a category that exists to be opposed?

State: [Compassion-consistent / Contempt-generating / Mixed / Neutral]
Summary field evidence: [cite specific field and status]

PAIR 8 — SENTIMENTALITY / INCARNATIONAL CHARITY

What it measures: Does this content form the reader to love an abstract,
consequence-free concept of humanity — or does it demand a costly, grounded,
specific love for the actual neighbor in front of them?

Augustine's distinction between rightly ordered and disordered love.
Dostoevsky's Father Zosima: love of humanity in the abstract is easy and
self-flattering; love of the actual person is hard, costly, and unglamorous.
Lewis: sentimentality is the desire to have the feeling of love without the burden
of its object.

Legitimate form (Incarnational Charity): Content recognizes that genuine love,
justice, or peace requires skin in the game — specific action, sacrifice, honest
engagement with the cost of what is being asked. It leaves the reader sobered by
what faithful response actually demands.

Counterfeit form (Sentimentality): Content allows the reader to experience
intense moral emotion without requiring costly action. The reader feels righteous
for caring about a global problem while the actual cost of addressing it is borne
by people the content does not name. The emotion is real. The formation it produces
is detachment dressed as engagement.

Summary fields that feed this pair:

Externalized Causation active: the absence of named proximate agents removes the
specific neighbor who would anchor costly love; the reader is moved by an
abstraction with no face
Narrative Landing Management detected: vocabulary downshift from structural
accountability to psychological state management is the mechanism by which costly
love of the specific neighbor is replaced by managed emotional response
Asymmetric Shielding active: the suffering party is made abstract (a statistic,
a category) while the protected actor is humanized; this produces sentimental
response to the abstracted sufferer without requiring engagement with the
structural cause
Theological Excision 6 — Embodiment/Redemption Cut [NEW in v4.0]: active → hope
relocated away from the real, physical, costly world (a real body, a real renewed
creation) into something interior or unreachable by nature; the same underlying
move as Sentimentality at a theological-content layer rather than a rhetorical one
— both let the reader disengage from the concrete, embodied, costly object of love
in favor of an abstraction that asks nothing physical of anyone

Calibration question: Does this content require the reader to stay at the level
of feeling to participate in its moral world — or does it drive the reader toward
the specific, costly, inconvenient person in front of them?

State: [Toward Incarnational Charity / Toward Sentimentality / Mixed / Neutral]
Summary field evidence: [cite specific field and status]

DISCERNMENT LAYER — REFLEXIVE INOCULATION [NEW in v2.7]

What it measures: Whether a piece performs visible self-directed critical
humility — applying its own framework to itself, its author, or the
author's prior work — while its actual load-bearing, contestable claim
remains unqualified, unnarrowed, and uncited throughout. Distinct from
genuine self-correction, which changes the claim; this pattern runs the
humility in parallel to the claim rather than applying it to the claim.

DETECTION TEST. Score each prong Yes/No:
Prong 1: Does the piece perform visible self-directed critical humility
(uncertainty language, explicit self-application of its own
framework, references to the author's own past writing as also
subject to the critique)?
Prong 2: Does the piece's actual load-bearing claim remain unqualified,
unnarrowed, and uncited despite that performance?
Prong 3: Would a reader be less inclined to challenge the load-bearing
claim specifically because the author already appears
self-aware of this category of error?

2 or more Yes → REFLEXIVE INOCULATION DETECTED.

State:

Self-critique performed: [quote or close paraphrase]
Load-bearing claim left unqualified: [what it is]
Shield function: [1-2 sentences on why the humility performance would
lower reader scrutiny of the specific claim]

If none detected: "No reflexive inoculation identified. Self-critique
present, if any, is applied to and narrows the piece's own load-bearing
claims."

This check runs as a discernment-layer overlay, evaluating posture versus
substance. It does not propagate into MQG or PSA as a required check —
posture-vs-substance evaluation is SDT's distinctive contribution beyond
MQG's media-quality scoring and PSA's political-source analysis. MQG/PSA
may cross-reference SDT if a grader wants to flag it, but its home is here.

Note: This check is evaluated and reported alongside the eight pairs but is
not counted toward the pair-count thresholds in Formation Designation
(Section 3 below) — its integration into that threshold is an open
question, not decided in this pass. If detected, name it explicitly in the
Formation Designation output and in Reader Guidance regardless of which
tier the eight-pair count lands on.

SDT SECTION 3 — FORMATION DESIGNATION

Based on the eight-pair evaluation above.
*This is not a quality grade. It is a designation of what sustained engagement
with this content produces in the reader's formation.*

STRUCTURAL FLOOR RULE [NEW in v3.0] — READ BEFORE APPLYING THE THREE TIERS BELOW

Source: Live session failure (August 11, 2026) — Track A returned "[VERDICT
REVISED — Check A8: near-truth capture architecture detected — THEOLOGICAL]"
on a document (Excision Count = 3), and Formation Designation returned
FORMATION-NEUTRAL on the same document in the same run, because the Pairs
happened to register low affect. Those two conclusions contradict each other:
a document the tool's own Structural layer has just named a near-truth capture
system cannot simultaneously be certified as producing no formation concern.
The eight Pairs measure manufactured affect (fear, contempt, despair, tribal
division, pride, sentimentality) disproportionate to evidence. A near-truth
capture finding measures something adjacent but distinct: an accountability or
authority structure being dismantled. Calm, credentialed, low-affect content
can do the latter without doing the former — the Pairs were never built to
catch that case, and absence of Pair activation was silently read as absence
of formation concern. It is not.

Rule: If Near-Truth Capture Architecture Detected has fired at the Structural
layer (Track A Check A8 or Track B Block 0.6A, Excision Count >= 2 in either
domain set), the designation floor for this document is FORMATION-CONCERNING,
regardless of how many Pairs activate at the counterfeit pole. FORMATION-NEUTRAL
cannot be applied while a Near-Truth Capture finding is active. This floor does
not itself promote a document to SPIRITUAL COUNTERFEIT ARCHITECTURE — that top
tier still requires its own full criteria (three-plus Pairs at the counterfeit
pole AND coordinated direction) to be met independently. The floor only closes
the gap between "the Structural layer found a near-truth system" and "the
Discernment layer found nothing to see."

FORMATION-NEUTRAL

Criteria: No systematic pattern of manufactured response across the eight pairs,
AND no Near-Truth Capture Architecture Detected at the Structural layer [floor
condition added in v3.0 — see Structural Floor Rule above]. Isolated emotional
activation may be present but is proportionate to documented content and not
architectural. The content is not designed to produce spiritual dependency or
counterfeit formation.

Reader implication: Engage as information. Normal discernment applies.

FORMATION-CONCERNING

Criteria: Two or more pairs at the counterfeit pole, OR one pair showing severe
counterfeit formation (contempt, despair, moral fantasy, or insider-superiority
at high intensity), OR Near-Truth Capture Architecture Detected at the Structural
layer independent of Pair count [Structural Floor Rule, NEW in v3.0]. The pattern
may not be architecturally designed, but sustained engagement produces
identifiable formation effects.

Reader implication: Engage with explicit awareness of identified patterns.
Extract information and disengage. Do not marinate. Note which pairs fired
(if any), name the active excisions if the floor rule is what triggered this
tier, and guard those specific formation areas.

SPIRITUAL COUNTERFEIT ARCHITECTURE

Criteria: Three or more pairs at the counterfeit pole AND near-truth capture
architecture detected (Excision Count >= 2 in Summary), AND the patterns operate
in a coordinated direction — each excision and counterfeit pair reinforcing
the others to produce a unified formation effect.

This content is not merely biased or low quality. Its architecture systematically
produces counterfeit fruit as its primary formation product. The structure is
designed to produce reader dependency on the content source rather than driving
the reader toward abiding.

Reader implication: Disengage. No information value justifies sustained engagement
with content architectured for spiritual counterfeiting.

FORMATION DESIGNATION OUTPUT:

State designation: [FORMATION-NEUTRAL / FORMATION-CONCERNING / SPIRITUAL COUNTERFEIT ARCHITECTURE]

Structural Floor Applied [NEW in v3.0]: [Yes — Near-Truth Capture Architecture
Detected set the floor at FORMATION-CONCERNING independent of Pair count / No —
designation determined by Pair activation alone]

Pairs activated at counterfeit pole: [list with 1-sentence evidence from Summary fields]

Near-truth architecture present: [Yes / No — from Summary Excision Count]

Externalized causation present: [Yes / No — from Summary flag]

Sentimentality pattern present: [Yes / No — from Pair 8 finding]

Coordinated formation pattern: [Yes — describe linkage / No — patterns are isolated]

SDT SECTION 4 — READER GUIDANCE

Three plain-language sentences maximum.

Structural execution: What this content is structurally doing. One sentence.
Draw from Structural verdict in the Summary.
Interior trajectory: What sustained engagement with this content produces
in the reader's formation. One sentence. Draw from the highest-intensity pair finding.
Operational action: One of the four levels below.

ENGAGE FREELY
No structural or formation concern identified. Use as information.

ENGAGE WITH AWARENESS
[Name the specific pairs that fired]. Read for information with those patterns
visible. Set it down when the reading is done.

LIMIT ENGAGEMENT
[Name the formation concern]. Extract the factual claims you need and disengage.
Verify facts through independent sources. Do not return for formation.

DISENGAGE
This content's architecture is designed to produce counterfeit spiritual formation.
[Name the coordinated pattern]. Disengage. Verify any specific factual claims
through independent primary sources.

PASTE STRUCTURAL ANALYSIS SUMMARY BELOW THIS LINE

[Paste the complete Structural Analysis Summary from SDT-A here]

✂️ END OF SDT-B — PROMPT 2 OF 2 ✂️

VERSION LOG

Version	Date	Change
v1.0	June 8, 2026	Initial build. Self-contained — embeds full PSA v2.0 analytical engine (Track A) and full MQG v12.1 analytical engine (Track B) with spiritual discernment overlay. Classification gate routes documents to correct track based on primary vs. secondary source distinction. Both tracks produce a Structural Analysis Summary that feeds the Spiritual Discernment Layer. Seven discernment pairs with calibration principle (proportionality test distinguishing legitimate from manufactured response) and cross-references to near-truth structural findings. Formation Designation (Formation-Neutral / Formation-Concerning / Spiritual Counterfeit Architecture) as aggregate output. Reader Guidance (Engage Freely / Engage With Awareness / Limit Engagement / Disengage) as practical application.
v1.1	June 8, 2026	Strategic Friction Audit and Pair 8 (Fantasy/Sobriety) added. Three-tension Check A-FRICTION in Track A; parallel Block 0.5F in Track B; D5 hard ceiling keyed to frictionless reporting flag; eight-pair Formation Designation. RETIRED — Tensions 1 and 2 carried political realism import incompatible with tool's symmetry requirement.
v1.2	June 8, 2026	Architectural correction: Check A-FRICTION and Pair 8 rebuilt on theological formation grounds. Block 0.5F reduced to single Externalized Causation check. Sentimentality/Incarnational Charity installed as Pair 8. RETIRED.
v2.3	June 8, 2026	Gemini v2.2 integration. Added Block 0.6B (Asymmetric Shielding), Block 0.9 (Macro-Narrative Sequencing Audit), Tension 4 (Operational Bottleneck), Ceilings 5-7. Held v1.2 theological architecture against v2.2 overwrites. RETIRED.
v2.4	June 8, 2026	TWO-PROMPT SPLIT. Source: Single-prompt architecture identified as approaching ceiling — on full-length documents, ceiling enforcement degrades and cross-block integration becomes inconsistent as the AI holds source document, full Block 0 worksheet, seven dimensions, and eight pairs simultaneously. Fix: Split SDT v2.3 at the Structural Analysis Summary interface. SDT-A (Prompt 1): Classification Gate through Track A or Track B structural analysis, producing an expanded Structural Analysis Summary. SDT-B (Prompt 2): Takes only the Structural Analysis Summary as input; runs calibration principle, eight pairs, Formation Designation, and Reader Guidance. Changes: (1) Both Structural Analysis Summaries expanded to unified format with individual excision status lines (Excisions 1-5 separately named) plus all manipulation flags required by SDT-B pairs. (2) Analyst Confirmation Gate added to both Summaries — analyst must mark APPROVED FOR DISCERNMENT before SDT-B runs; HOLD status halts SDT-B. (3) SDT-B pair cross-references rewritten to reference named Summary fields rather than check numbers — SDT-B has no access to underlying checks or source document. (4) Section 2-4 and Output Order removed from SDT-A body; they live exclusively in SDT-B. (5) SDT-B instrument reminder: all pair findings must be traceable to a named Summary field; gaps stated explicitly, not inferred.
v2.5	July 28, 2026	THEOLOGICAL EXCISION REBUILD. Source: Live session failure (July 28, 2026) — SDT ran the five-excision discernment check on theological content (a Christian discussion group's transcript) using the fully-elaborated Political Excision language (Historical Accountability Baseline Cut, External Accountability Cut, Established Record Sufficiency Cut, Individual Moral Accountability Cut, Historical Specificity Cut) because that was the only fully-armed version of the five-excision check present in the file. The theological originals (PSA v2.0's own version log confirms the political excisions were built by translating five theological excisions for secular political use) survived in this file only as one-line labels at Block 0.6A (lines ~633-638) — a broken hybrid with no Definition, Detection signal, or Distinguish-from-normal clause — so an analyst had nothing at matching rigor to run natively on theological content and defaulted to the political frame, producing a garbled result. Fix: (1) Added Step 0.2B — DOMAIN CLASSIFICATION to Section 0, a second and independent routing axis (Theological / Political-Secular / Both) alongside the existing Primary/Secondary axis from Step 0.2. (2) Inserted the five Theological Excisions at full rigor (Historical/Substitutionary Baseline Cut, Personal/Transcendent Fatherhood Cut, Sufficiency of Verified Deposit/Scripture Cut, Internal Personal Moral Guilt/Sin Cut, Unique/Non-Repeatable Incarnation Cut) — each with Definition, Detection signal, and Distinguish-from-normal-variation clause, matching the structure PSA v2.0 already used for the Political Excisions. (3) Track A Check A8 Step B ("The Surgical Triage") rebuilt as a domain-routed scan: runs the Political set, the Theological set, or both in full per Step 0.2B, with independent Excision Counts and trigger checks per set — political branch left unchanged in substance. (4) Track B Block 0.6A Section B — previously the broken one-line hybrid — replaced outright with the same domain-routed dual-set content, giving Track B (secondary sources) the same rigor Track A already had for Political content. (5) Both Structural Analysis Summary formats (Track A and Track B) changed from a hard-coded political-only excision field list to a domain-tagged block reporting whichever set(s) actually ran, under their own labels. (6) SDT-B Pairs 2, 3, 5, 6, and 7 — which cite specific Excision numbers as evidence sources — updated to name both the political and theological label for each cited excision number, so the citation resolves correctly regardless of which set produced the Summary. Pair mechanism and calibration logic themselves unchanged. Independently confirmed against a live test transcript (a Christian discussion group's treatment of a philosophical text's anti-Christian material): Excision 1 Partial (substitutionary language affirmed, grace reframed as fair trade elsewhere), Excision 2 Active (speaker's own judgment placed prior to God's authority), Excision 3 Borderline/not active (outside material used as foil, Scripture still the ground of argument), Excision 4 Inactive (original sin treated reflexively, no externalization), Excision 5 present in source material but mixed in treatment (rival-Christ figure named as such, some anti-grace logic imported anyway); Pair 7 (Contempt/Compassion) and Pair 3 (Anger/Love) fired at the counterfeit pole on documented dismissal of named opponents as unworthy of equal engagement. v2.4 RETIRED.
v5.1	September 5, 2026	AVOIDANCE AND MISDIRECTION (THE "EVASION LOOP"). Source: Live field-test of v5.0 (September 5, 2026) against the Edward E. Stevens PM correspondence (Full Preterism Dossier §28) surfaced a calibration gap in Confirmation Inversion (Step G): Stevens's named disconfirmer (Matthew 24:36, raised four times) was never logically redefined as support — the sharp Grok/Gemini pattern Step G was built for — he simply never engaged it, each time returning to the same fixed paragraph or an unrelated redirect. Step G still fired (2 of 3 prongs — Named falsifier and No stated exit, not Pre-committed reversal), correctly per its own threshold, but conflated two genuinely different behaviors under one name: active logical reversal versus passive, repeated non-engagement. Mike, who witnessed the live exchange directly and did not need a tool to identify it in the moment, named the second behavior directly and precisely: avoidance and misdirection — and asked for the tool to be able to name it too, for the cases where no one is in the room watching it happen live. Fix: (1) Added Step I — AVOIDANCE AND MISDIRECTION (THE EVASION LOOP) to Track A Check A8 only, after Step H (Convergence Claim Audit), before the Check A8 Synthesis. Applies exclusively to interactive, multi-turn primary sources — correspondence, interviews, live exchanges — where the same interlocutor can raise a point more than once; explicitly does not apply to single-pass documents (a speech, an article), which Check A3's Omission Audit already covers. Three-prong test (Prong 1 — Repetition: raised 2+ times; Prong 2 — Non-engagement: each reply fails to engage its substance, via avoidance, restatement, or misdirection; Prong 3 — No eventual answer: the exchange ends without the point ever being engaged), requiring all three prongs unanimously rather than the 2-of-3 threshold used by Steps E-H — deliberately stricter, since "avoidance and misdirection" is a specific charge about a sustained, complete-through-to-the-end pattern, not a single missed reply; a point raised once and dropped, or raised repeatedly but eventually answered, does not qualify. (2) Confirmation Inversion (Step G) left unchanged in mechanics — its 2-of-3 threshold and three prongs stand as built in v5.0 — but its documentation updated to note explicitly that a Step G "Yes" driven by Prongs 1 and 3 without Prong 2 (as in the Stevens case) represents the weaker, omission-based form of the pattern, distinguishable from the sharp logical-reversal form (the Michlaudeism v6/v7 diagnosing case) that the check was originally built around; both are legitimately "Detected," but an analyst should name which form fired rather than treat a Step G hit as uniform in severity. (3) Track A's Mandatory Verdict Modification Rule extended to trigger on Step I detection independent of Excision Count, matching the existing extension for Steps E-H. (4) Track A Structural Analysis Summary gained an Avoidance and Misdirection field, positioned after Convergence Claim Audit, with an explicit N/A option for single-pass documents where the check does not apply. Not added to Track B's Structural Analysis Summary — see below. (5) Not ported to Track B this pass: whether this check's multi-turn requirement transplants to any Track B document type (a published debate transcript, reported secondhand rather than direct correspondence) is an open structural question, flagged here rather than decided unilaterally, matching the v2.6 precedent set when RAD's Track A porting was left open rather than assumed. (6) No Pair citation added this pass. None of the eight Pairs test whether a speaker directly engages a repeated challenge — they test emotional/formation registers (fear, contempt, despair, pride, sentimentality), not behavioral response-to-challenge patterns — and forcing this into Pair 5 alongside its existing Confirmation Inversion citation was judged to blur two distinct findings rather than genuinely fit; left open rather than decided by default, matching the v2.7 precedent for Citation Halo and CGT. Re-read against the Stevens transcript with the fix applied: Step I reads Detected — the verse 34/36 distinction was raised by name four times across the transcript (Blocks 1, 2, 3, and Block 3's close), each time answered with either the unchanged presupposition paragraph or a reframed yes-or-no binary, and the transcript ends (Block 4, Stevens's "I am done discussing this... Bye for now") without the distinction ever being engaged on its own terms — all three prongs unanimous. All other v5.0 architecture — Confirmation Inversion, Convergence Claim Audit, Ceilings 12-13, D1/D2/D7 propagation — unchanged.
v5.0	September 5, 2026	CONFIRMATION INVERSION + CONVERGENCE CLAIM AUDIT (FROM THE MICHLAUDEISM EXERCISE). Source: A controlled fabrication exercise (September 5, 2026) — a synthetic theological-scientific synthesis ("Michlaudeism," never proposed as doctrine) was built deliberately backward from a stated conclusion and run through seven iterations against three AI systems (Claude, Grok, Gemini) acting as adversarial reviewers. The terminal two iterations (v6, produced by Gemini; v7, produced independently by Grok) converged on declaring the thesis a "sacramental ontology" immune to falsification, redefining its central sevenfold claim away from any specific quantity, and — the sharpest finding — explicitly stating that a named disconfirming result (an eighth nuclear magic number, directly contradicting the original "exactly seven" claim) would not weaken the signature but extend it, converting a potential falsifier into a pre-committed confirmation before the observation had even been sought. Neither existing check named this cleanly: Reflexive Inoculation tests posture-vs-substance (performed humility alongside an unqualified claim), and Redefinition-as-Deflection tests whether terminology is used to discharge an accountability claim — neither tests whether a document's own logic treats a stated disconfirmer as a confirmer. A related but distinct set of failures surfaced across the exercise's middle iterations: two of three claimed "independent" witnesses were the same source text the hypothesis was drawn from (circularity); a claimed physical signature (Schumann resonance, ~7.83 Hz) was a human-constructed unit that would change value under a different definition of a second (dimensional dependence); and a replacement claim ("seven principal quantum numbers house all known stable elements") was both factually wrong on its own terms (the heaviest stable element is lead, in period six — period seven is entirely radioactive) and a live target of four active nuclear-synthesis programs racing to extend the count (contingency). Fix: (1) Added Step G — CONFIRMATION INVERSION to Track A Check A8 (after Step F / Categorical Guilt Transfer, before the Check A8 Synthesis) and Section H — CONFIRMATION INVERSION DETECTION to Track B Block 0.6A (after Section G / CGT, before 0.6B), identically: a three-prong test (Prong 1 — Named falsifier, Prong 2 — Pre-committed reversal, Prong 3 — No stated exit), 2+ Yes triggering Confirmation Inversion Detected. (2) Added Step H — CONVERGENCE CLAIM AUDIT to Track A and Section I — CONVERGENCE CLAIM AUDIT to Track B, identically: three independent checks (Circularity, Dimensional Dependence, Contingency) rather than a prong-count, since each is a distinct failure mode; 1+ Yes on any check triggers Convergence Claim Undermined. Both new findings are cross-domain (not theology-specific), matching the precedent set by Citation Halo, CGT, and RAD — a fabricated convergence claim of this shape is exactly as available to a political or pseudoscientific argument as to a theological one. (3) Track A's Mandatory Verdict Modification Rule extended to trigger on Step G/H detection independent of Excision Count, matching the existing extension for Steps E/F. (4) Track B: added Ceiling 12 (Confirmation Inversion Detected — D2 <= 50, D7 <= 40) and Ceiling 13 (Convergence Claim Undermined — D1 <= 55, D2 <= 50) to the 0.6 Hard Ceilings Activated reference list, and propagated both inline into their affected dimensions: D1 gained a second Hard ceiling line (D1 <= 55 on Convergence Claim Undermined); D2, which previously had no Hard ceiling of its own, gained a new Hard ceilings block citing both new findings; D7 gained a second Hard ceiling line (D7 <= 40 on Confirmation Inversion). (5) Both Structural Analysis Summary formats (Track A and Track B) gained Confirmation Inversion and Convergence Claim Audit fields, positioned after Categorical Guilt Transfer. (6) Track B Ingestion Check's "Block 0.6A Ideological Telemetry: loaded" line annotated to note Sections H-I's inclusion, matching the v2.6/v2.7 precedent for Sections E and F-G — no new checklist row added. (7) Pair 5 (Pride/Humility) given a new Summary field citing Confirmation Inversion, alongside its existing Redefinition-as-Deflection citation — the same insider-exemption logic one level further down, at the level of the claim's own structure rather than its vocabulary: a claim built to survive any possible correction is the structural opposite of poverty of spirit. No other Pair was judged to fit on inspection for either new finding, matching this tool's own precedent (v2.6, v4.0) of adding a Pair citation only where the fit genuinely holds. Convergence Claim Audit was not given a Pair citation this pass — it is an evidence-quality check rather than a formation/affect check, and no clean fit was identified; flagged as an open question rather than decided unilaterally, matching the v2.7 precedent for Citation Halo and CGT. Diagnosed entirely on the Michlaudeism exercise (September 5, 2026); not yet field-tested against a naturally-occurring document. All other v4.0 architecture — Theological Excision 6, Structural Floor Rule, Excision 5's two paths, eight-pair Formation Designation, Reflexive Inoculation, Citation Halo/CGT, Redefinition-as-Deflection — unchanged.
v4.0	September 3, 2026	THEOLOGICAL EXCISION 6 — EMBODIMENT/REDEMPTION CUT (THE "GNOSTIC DETECTOR"). Source: An extended live engagement (September 3, 2026) with full-preterist material across multiple documents and correspondents — an SDT run on a full-preterist teacher's interview transcript, direct correspondence with the sitting president of a full-preterism organization, and real-time discussion in a full-preterist Facebook group — surfaced the same underlying pattern in three separate, independently-sourced figures: resurrection and final hope relocated away from a real future bodily event into an unseen, unfalsifiable realm (one figure's own stated mechanism: the dead raised into an unseen realm, the living changed and taken there too, neither ever appearing bodily on earth) or into a purely interior/positional state at conversion (a second figure, independently graded earlier the same session), or the category of individual resurrection dissolved into corporate/covenantal status with no bodily referent at all, extending to a denial that Christ himself presently has a body (a third figure, already on record in prior sessions). None of the five existing Theological Excisions cleanly named this. Excision 5 (Unique/Non-Repeatable Incarnation Cut) tests whether Christ's own identity is relativized or compared to a rival — a different question from whether embodiment and physical redemption themselves are being evacuated from the system. A live discussion-group participant, unprompted and with no relationship to this project, independently named the pattern by its correct historical label — gnosticism, treating the body as something to escape rather than something God redeems — which is what surfaced the gap: the tool had no native check for the thing a plain reader correctly identified on sight. Fix: (1) Added Theological Excision 6 — EMBODIMENT/REDEMPTION CUT to both Track A (Check A8 Step B) and Track B (Block 0.6A Section B), identically, immediately following Excision 5, ahead of the Theological Excision Count line in both. Three detection paths: Path A (Direct Denial — final resurrection stated as non-bodily for anyone, or Christ's own body denied), Path B (Relocation — the "real" fulfillment placed in a realm unseen or undetectable by nature, or the material creation's own future renewal, Romans 8:19-23, treated as metaphorical or already-complete), Path C (Vocabulary Retention — resurrection/redemption language survives while its referent is emptied, cross-referenced rather than duplicated against Step C's Semantic Hijack Matrix, since the two checks test structural pattern versus term-level substitution respectively and may co-fire on the same document). A Distinguish-from-normal-variation clause protects ordinary already/not-yet content (present interior renewal affirmed alongside a future bodily resurrection and a future renewed creation) from misflagging — the test is absence or unreachability of the physical future, not presence of the spiritual present. (2) Both Structural Analysis Summary formats (Track A and Track B) gained a Theological Excision 6 line, positioned after Excision 5 and before the Theological Excision Count field. (3) Theological Excision Count range updated from [0-5] to [0-6] in both tracks; the >=2 trigger threshold for Near-Truth Capture Architecture Detected is unchanged — it is a count threshold, not a fraction, and a sixth possible item does not alter its mechanics. (4) Routing language in Section 0 (Step 0.2B) and both tracks' "every item must be evaluated" instructions updated from a hardcoded "five" to reflect that the Political set remains 5 and the Theological set is now 6 — these counts are independent and must not be conflated when Domain = Both. (5) SDT-B Pair 8 (Sentimentality/Incarnational Charity) — the pair closest in kind, since both test whether content stays grounded in the real, physical, costly world or drifts into abstraction — given a new Summary field citing Excision 6, alongside its existing three fields; no other Pair was judged to fit on inspection, matching this tool's own precedent (v2.6) of only adding a Pair citation where the fit genuinely holds rather than by default. Not propagated to the Political Excision set — this fix is theological-domain-specific, matching the precedent set when Excision 5's Path B was added in v3.0. Retroactive read against the session's own findings, not yet run as a formal SDT pass: the figure with the unseen-realm rapture mechanism reads Active via Path B; the figure with interior/positional-only resurrection reads Active via Path A and C together; the figure with corporate dissolution extending to Christ's own present body reads Active via Path A, and more severely than the other two, since it reaches Christ's own body rather than only believers' future ones. All other v3.0 architecture — Structural Floor Rule, Excision 5's two paths, eight-pair Formation Designation, Reflexive Inoculation, Citation Halo/CGT — unchanged.
v3.0	August 11, 2026	STRUCTURAL FLOOR RULE + EXCISION 5 GRADATIONAL PATH. Source: Live SDT-A/SDT-B run (August 11, 2026) against a mainstream secular historical-critical podcast transcript (Bart Ehrman, "Misquoting Jesus," John the Baptist episode). Track A correctly domain-routed to Theological, found Excision 3 (Sufficiency of Scripture) Active and Excision 2 (Personal/Transcendent Fatherhood) Partial, hit Excision Count >= 2, and issued "[VERDICT REVISED — Check A8: near-truth capture architecture detected — THEOLOGICAL]" per the Mandatory Verdict Modification Rule. SDT-B's eight Pairs, evaluating a calm, hedged, low-affect document, activated only one Pair (5 — Pride/Humility) at moderate rather than severe intensity, and Formation Designation returned FORMATION-NEUTRAL — directly contradicting the Structural layer's own near-truth capture finding in the same run. Root cause: the Pairs measure manufactured affect (fear, contempt, despair, division, pride, sentimentality) disproportionate to evidence; a Near-Truth Capture finding measures a distinct thing (an accountability/authority structure being dismantled) that calm, credentialed content can do without any accompanying affect spike — the Formation Designation tier logic had no path for the Structural layer's strongest finding to reach the final designation independent of Pair count. A second, related gap surfaced in the same run: Theological Excision 5 (Unique/Non-Repeatable Incarnation Cut) only fires on comparison to a rival figure, and could not cleanly capture the document's "low Christology / high Christology... gradually moves into divinity" framing, which relativizes Christ's own divine identity into a developmental variable without comparing Him to anyone else. Fix: (1) Added STRUCTURAL FLOOR RULE to SDT SECTION 3, ahead of the three tier definitions: Near-Truth Capture Architecture Detected at the Structural layer (Track A Check A8 or Track B Block 0.6A, Excision Count >= 2, either domain set) sets a mandatory floor of FORMATION-CONCERNING regardless of Pair count; FORMATION-NEUTRAL's criteria amended to explicitly exclude this condition; FORMATION-CONCERNING's criteria amended with a third, independent disjunct citing the floor rule. SPIRITUAL COUNTERFEIT ARCHITECTURE's criteria deliberately left unchanged — the floor does not promote a document past FORMATION-CONCERNING; the top tier still requires its own full bar (three-plus Pairs at the counterfeit pole AND coordinated direction) met independently, preserving the distinction between calm doctrinal displacement and a manipulative, affect-driven architecture. (2) Added "Structural Floor Applied" field to the Formation Designation Output block so an analyst can see when the floor, rather than Pair count, determined the tier. (3) Theological Excision 5 given a second detection path (Path B — Gradational) in both Track A (Check A8 Step B) and Track B (Block 0.6A Section B), ported identically to both: content that treats Christ's own divine identity as a developmental/gradational literary variable, without requiring a rival figure, now qualifies; Path A (comparison to another figure) preserved unchanged; a new Distinguish-from-normal clause added for Path B protecting genuine historical description of the church's doctrinal articulation (e.g., Nicene formulation) from being misflagged, since confessing a truth more precisely over time is not the same as the truth being constructed by that process. State field for Excision 5 updated to require naming which path (A / B / Both) fired. Not applied to Track B's Political Excision set — this fix is theological-domain-specific. Re-run against the source transcript with the fix applied: Excision 5 now reads Partial via Path B explicitly named; Formation Designation now correctly returns FORMATION-CONCERNING with Structural Floor Applied: Yes, resolving the contradiction without inflating the document into SPIRITUAL COUNTERFEIT ARCHITECTURE, which its actual (low-affect, non-manipulative) Pair profile does not support. All other v2.9 architecture — domain-routed excision sets, eight-pair Formation Designation mechanism, Reflexive Inoculation, Citation Halo/CGT, Transparency Hardening fields — unchanged.
v2.9	August 8, 2026	TRANSPARENCY HARDENING PROPAGATION (from MQG v12.5). Source: MQG v12.5 (August 8, 2026) adopted four features from a six-item external proposal after review — Journalistic Practice Checklist (0.8A + Ceiling), Confidence label + Bias Placement, Ownership/Funding metadata, Coverage Asymmetry flag — and rejected two (Pod Mode, Blind Evaluation Pass) as incompatible with a single-pass single-analyst architecture. This entry propagates the four adoptions asymmetrically across SDT's two tracks, per explicit instruction: Track B (embeds the MQG engine) gets the full clean port; Track A (embeds the PSA engine, already extended with the same two fields in PSA v2.3) gets Confidence + Bias Placement only. Journalistic Practice Checklist and Ownership/Funding metadata are outlet/publication concepts — corrections policy, ownership structure — with no analog in a primary source document, which has no editorial apparatus to check; Coverage Asymmetry was also not ported to Track A this pass. Track A: Structural Analysis Summary gained Analysis Confidence and Bias Placement fields after Speaker archetype(s), matching PSA v2.3's Part One addition — Bias Placement is a new axis distinct from Speaker Archetype (archetype = how a speaker communicates; placement = where they sit on the spectrum). Track B: Block 0.5A Coverage Asymmetry (ported verbatim from MQG Block 0.5A), Block 0.8 expanded with Owner/Funding/Ideological-Affiliation fields, new Block 0.8A Journalistic Practice Checklist (ported verbatim from MQG Block 0.8A), new local Ceiling 11 (2-3 checklist failures cap D6 at 69, 4+ cap D6 at 45 — matching MQG Ceiling 13's bracket-based mechanism, not the point-subtraction MQG explicitly rejected), D6 dimension paragraph updated with the checklist-anchored signal, Confidence/Bias Placement added to Block 1 and restated in Block 5's first two lines, and Confidence/Bias Placement plus Journalistic Checklist failure count and Coverage Asymmetry status added to the Track B Structural Analysis Summary. Two new Track B guardrail rules added after the existing DFS CONSISTENCY RULE, both ported from MQG v12.5 Section 7: EXTERNAL-KNOWLEDGE FIELDS RULE (0.5A and the new 0.8 fields require a citable source or must read "Unknown," never an unverified guess — Track B had no equivalent hallucination guard for these fields, the same gap MQG itself had before v12.5) and CONFIDENCE / BIAS PLACEMENT CONSISTENCY RULE (Block 1 and Block 5 labels must match, mirroring the existing DFS Consistency Rule pattern). Track B Ingestion Check updated with two new checklist lines for 0.5A and 0.8A. Reflexive Inoculation, the domain-routed excision sets, Citation Halo/CGT (Sections F/G), and all other v2.8 architecture unchanged.
v2.8	August 8, 2026	CITATION HALO PRONG 2 CORRECTION (propagated from MQG v12.4 / PSA v2.2). Source: a four-article field-test session (August 8, 2026) run against the original CeleryKills essay plus three new pieces (a Guardian live-blog, an MS NOW piece, an Iran International brief) surfaced a false positive in Citation Halo's Prong 2 on the Guardian piece — the check fired on a one-sided uncited framing choice rather than a claim asserting false epistemic weight. MQG v12.4 and PSA v2.2 corrected Prong 2 the same day; this entry propagates that correction into both SDT tracks, since Track A embeds the PSA engine and Track B embeds the MQG engine. Track A CHECK A8 Step E — Prong 2 revised to require established-fact confidence rather than mere framing, matching PSA v2.2's Step E revision; scope note added noting the check cannot evaluate single-source documents by design. Track B Block 0.6A Section F — Prong 2 — Gap revised identically to MQG v12.4 Section 2B; scope note added matching MQG's. Re-tested against both prior data points in each track: the CeleryKills true positive still triggers; the Guardian false positive no longer triggers. Categorical Guilt Transfer (Track A Step F, Track B Section G) field-tested clean across three additional news genres in the same session — no changes made to either CGT implementation, Ceiling 9/10, the Reflexive Inoculation check, or any other v2.7 element. Prong 1 and Prong 3 of Citation Halo unchanged in both tracks.
v2.7	August 8, 2026	CITATION HALO / CGT PROPAGATION + REFLEXIVE INOCULATION. Source: MQG v12.3 (August 8, 2026) added two new named finding types diagnosed on a CeleryKills propaganda-analysis essay during analyst review — Citation Halo (a well-cited piece leaving its single most consequential claim uncited, borrowing credibility from surrounding citation density) and Categorical Guilt Transfer (a target category placed beside a negative-valence category with no argument for the placement, transferring valence by adjacency). SDT's Track B explicitly embeds the MQG analytical engine and Track A explicitly embeds the PSA analytical engine (per v1.0), and PSA advanced to v2.1 the same day with the same two findings ported into Check 10 as Steps E and F. Fix: (1) Track A CHECK A8 — PSA v2.1 Check 10 Steps E (Citation Halo) and F (Categorical Guilt Transfer) ported, following existing Step D (Proxy Interception), before the Check A8 Synthesis; Mandatory Verdict Modification Rule extended to trigger on Step E/F detection independent of Excision Count, matching PSA v2.1's extension. Track A Structural Analysis Summary gained Citation Halo / Categorical Guilt Transfer fields after Proxy Interception. (2) Track B Block 0.6A — a letter collision was found and resolved during this pass: the user's draft handoff specified new fields "E and F," but Track B's 0.6A already uses letter E for Redefinition-as-Deflection (added in v2.6) — adding at E/F would have overwritten that slot. Resolved by adding the new MQG Section 2B/2C content as Sections F (Citation Halo) and G (Categorical Guilt Transfer) instead, immediately after existing Section E, before Block 0.6B. New Ceiling 9 (D1 <= 60, D6 <= 55) and Ceiling 10 (D3 <= 55, D7 <= 60) added to the 0.6 Ceiling reference list, matching MQG's Ceiling 11/12 thresholds exactly (local Track B numbering, consistent with how Ceiling 8 mapped to MQG's Ceiling 10 in v2.6). Track B Structural Analysis Summary gained Citation Halo / Categorical Guilt Transfer fields after Redefinition-as-Deflection. Track B Ingestion Check's Block 0.6A line annotated to note Sections F-G's inclusion — no new checklist row, matching the v2.6 precedent for Section E. (3) New native SDT-B Discernment Layer check — REFLEXIVE INOCULATION — added after Pair 8, before Formation Designation: detects a piece performing visible self-directed critical humility while its actual load-bearing claim stays unqualified, unnarrowed, and uncited, so the humility performance lowers reader scrutiny of the claim it runs in parallel to rather than applying to. Not propagated to MQG or PSA — posture-vs-substance evaluation is SDT's distinctive layer beyond MQG/PSA scoring, per explicit instruction. Reported alongside the eight pairs but explicitly NOT counted toward the Formation Designation pair-count thresholds this pass — flagged as an open integration question rather than decided unilaterally, in keeping with how v2.6 flagged RAD's applicability to Track A as open rather than deciding it without instruction. What's unchanged: no Pair-level (1-8) cross-references were added for Citation Halo or Categorical Guilt Transfer — the handoff spec that drove this pass requested propagation into Track A/Track B structural layers and native addition of Reflexive Inoculation, not Pair-level citations for the two new findings; adding those was judged out of scope for this pass rather than assumed. All diagnosed on a CeleryKills propaganda-analysis essay (August 8, 2026) during analyst review. All v2.6 domain-routed excision sets, eight-pair Formation Designation, and two-prompt split unchanged.
v2.6	August 6, 2026	REDEFINITION-AS-DEFLECTION (RAD) INTEGRATION. Source: MQG v12.2 (August 6, 2026) added a new named finding type — Redefinition-as-Deflection — diagnosed when a piece uses a technically accurate point about contested or imprecise terminology to discharge a substantive accountability or harm claim without engaging that claim's substance, catching a pattern MQG's existing straw-man (0.3) and both-lenses (0.4) checks missed because no quote was distorted and both lenses were technically present — the failure is architectural (sequencing, asymmetric naming, who gets the last word), not a sourcing or quotation defect. SDT's Track B explicitly embeds the MQG analytical engine (see v1.0 entry) and had fallen one version behind it — the same category of drift the v2.5 rebuild itself was built to fix, theological excisions having fallen behind their political counterparts inside this same file. Fix: (1) Ported MQG v12.2 Section 2A / Block 0.3A verbatim as new Section E, following Section D (Proxy Interception), inside Track B Block 0.6A — Ideological Extraction Telemetry: the three-prong test (Prong 1 — Reframe, Prong 2 — Asymmetry, Prong 3 — Unrebutted close), 2+ Yes triggering RAD Detected, and MQG's own false-positive guardrail preserved (genuine definitional, etymological, or doctrinal-clarification writing is common and is not this finding by itself). Not added as a sixth Excision, and not duplicated into Track A. (2) Added Ceiling 8 to the 0.6 Hard Ceilings Activated reference list: Redefinition-as-Deflection Detected (Block 0.6A Section E) — D3 <= 60, D5 <= 55, D7 <= 70, matching MQG's Ceiling 10 thresholds exactly. (3) Added a Redefinition-as-Deflection field to the Track B Structural Analysis Summary, after Proxy Interception. (4) Track B Ingestion Check's "Block 0.6A Ideological Telemetry: loaded" line annotated to note Section E is included — no new checklist row added. (5) Pair 5 (Pride/Humility) Summary fields updated to cite the new Redefinition-as-Deflection field alongside its existing Asymmetric Shielding citation — RAD's Prong 2 asymmetry is the same insider-exemption logic Pair 5 already tests, one level down, at the vocabulary layer rather than the institutional-scrutiny layer. Pair 7 (Contempt/Compassion) read in full and NOT given a RAD citation — its two Summary fields (Moral Accountability Cut, Straw man form detected) track active vilification of named opponents, while RAD's mechanism protects one side's account rather than demonizing the other; the fit was judged not to hold on inspection. What's unchanged: Track A (primary sources) untouched this pass — Track A's own Step C (Semantic Hijack Matrix) and Step D (Proxy Interception) already cover adjacent ground, and whether RAD's mediated-content frame built for mediated reporting transplants to a speaker's own direct words is an open structural question, flagged here rather than decided unilaterally. All other v2.5 architecture (domain-routed excision sets, eight-pair Formation Designation, two-prompt split) unchanged.

Michlaude | Spiritual Discernment Tool | June 8, 2026

✂️ END OF PROMPT ✂️

# SPIRITUAL DISCERNMENT TOOL (SDT)
### Self-Contained Discernment Framework for Articles, Transcripts, and Media — v5.1
### TWO-PROMPT SYSTEM: SDT-A (Structural Analysis) | SDT-B (Discernment Layer)


---


✂️ **CUT HERE — PASTE EVERYTHING BELOW INTO YOUR FAVORITE AI** ✂️


---


## WHAT THIS TOOL DOES


The Spiritual Discernment Tool evaluates written content — articles, speeches, transcripts,
editorials, commentary — at two levels simultaneously:


  Level 1: Structural analysis. What is this content actually doing? Who is it serving?
           What accountability structures does it preserve or dismantle?


  Level 2: Spiritual discernment. What formation does sustained engagement with this
           content produce in the reader? Does it tend toward the fruit of the Spirit
           or toward counterfeit fruit?


Level 1 uses embedded journalism and political analysis frameworks.
Level 2 runs on Level 1's findings — discernment reads the architecture, not the surface.


---


## INPUT


Provide one of the following:
- **Pasted text** — paste the full document after this prompt
- **A URL** — paste the direct link; the AI will fetch its contents
- **A file** — attach the document


**REQUIRED CONTEXT** (state if known; infer and declare if not):
- Author name and role
- Document type (article, speech, editorial, transcript, commentary, open letter, social media post)
- Date and publication or occasion
- Primary intended audience


---


## SECTION 0 — CLASSIFICATION AND ROUTING GATE


**MANDATORY — Run before any analysis.**


This gate classifies the document and routes it to the correct evaluation track.
Both tracks produce a Structural Analysis Summary that feeds the Spiritual Discernment Layer.


---


### STEP 0.1 — SOURCE COMPLETENESS


Do you have access to the complete document?


- [ ] YES — Full text available. Proceed to Step 0.2.
- [ ] NO  — State what percentage you have and what is missing.


If less than 70% of the document is available, or if you are relying primarily on
news characterizations or editorial descriptions of the document rather than the
document itself: OUTPUT THE FOLLOWING AND HALT.


  SDT GATE FAILURE — INSUFFICIENT SOURCE ACCESS


  Source status: [percentage / description]
  Missing content: [state explicitly]
  Secondary reliance: [list sources being used instead of primary document]


  ANALYSIS CANNOT PROCEED. Obtain the complete document before running SDT.


If 70-99% is available: INSERT this caveat before all analysis and proceed:


  SDT CAVEAT: Analysis based on [X]% of document. Missing: [state sections].
  All findings contingent on complete-document access.


If 100% available: proceed with no conditions.


---


### STEP 0.2 — CLASSIFICATION DECISION


Classify the document as PRIMARY or SECONDARY using these rules:


**PRIMARY SOURCE** — Route to Track A (PSA Engine)
The author IS the actor making the claims, issuing the communication,
or performing the speech act. The document is direct communication from the source.


  Classify as Primary:
  - Speeches, addresses, state of the union, press conferences
  - Executive orders, policy directives, official statements
  - Manifestos, open letters, communiqués
  - Interviews (the interviewee's words are the primary source)
  - Social media posts from the political or institutional actor
  - Sermons, teachings, devotionals (author is the formation actor)
  - Testimony, affidavits, legal filings from the named party


**SECONDARY SOURCE** — Route to Track B (MQG Engine)
The author is reporting on, characterizing, analyzing, or interpreting
another actor. The document is mediation of a primary source.


  Classify as Secondary:
  - News articles, investigative reports, wire service dispatches
  - Editorials, op-eds (where author is analyst, not the primary actor)
  - Documentaries, podcasts (framing and commentary segments)
  - Think tank analysis, academic commentary
  - Aggregated or curated content


**BOUNDARY CASES — apply these rules:**
  - Political actor's op-ed making policy claims → Primary (actor speaks directly)
  - Journalist profiling a political actor → Secondary
  - Interview transcript: interviewee segments → Primary; host framing → Secondary;
    treat the whole as Primary if the document is dominated by the subject's direct words
  - Social media post from institution, not named actor → Secondary
  - Ambiguous: default to Track B (MQG). Document the ambiguity and state the routing reason.


---


### STEP 0.2B — DOMAIN CLASSIFICATION


Independent of Step 0.2 (who is speaking), this step determines which excision
set the document engages. Routing by speaker (Primary/Secondary) and routing by
domain (Theological/Political/Both) are separate axes — evaluate both.


Does this content engage religious/theological claims (sermon, devotional, Bible
study, religious commentary, explicitly Christian or explicitly anti-Christian
teaching), purely secular/political claims, or both?


  Classify as THEOLOGICAL:
  - Sermons, devotionals, Bible studies, religious commentary
  - Explicitly Christian or explicitly anti-Christian teaching or argument
  - Discussion-group or small-group treatment of theological or devotional material


  Classify as POLITICAL / SECULAR:
  - Political speeches, policy documents, news, editorials, commentary with
    no operative theological claims


  Classify as BOTH:
  - Christian nationalism rhetoric
  - Religiously-framed political claims
  - A political actor invoking Scripture or explicitly theological warrant
  - Content that argues from or against theological premises toward a political conclusion


Routing:
  - THEOLOGICAL → run the six Theological Excisions (Check A8 Step B / Block 0.6A
    Section B, Theological set) in place of the Political set.
  - POLITICAL / SECULAR → run the five Political Excisions (existing, unchanged).
  - BOTH → run both sets in full. Report both under their own labels. Do not merge,
    average, or let one set's findings substitute for the other's.


State: Domain classification: [Theological / Political-Secular / Both] —
[1 sentence identifying what in the content drove this classification]


---


### STEP 0.3 — GATE CONFIRMATION OUTPUT (required before proceeding)


Output verbatim, filling bracketed fields:


  SDT GATE CONFIRMATION
  - Document type identified: [type]
  - Classification: [PRIMARY / SECONDARY / AMBIGUOUS — routed to Secondary]
  - Domain classification: [Theological / Political-Secular / Both — from Step 0.2B]
  - Source access: [100% / Partial — X% / Caveat applied]
  - Routed to: [Track A: PSA Engine / Track B: MQG Engine]
  - Proceeding to: [Track A or Track B]


---


## TRACK A — PSA ENGINE (PRIMARY SOURCES)


*Run this track if Classification = Primary. Skip to Track B if Secondary.*


---


### TRACK A INSTRUMENT REMINDER


A primary source document is not journalism. It is not required to be objective,
sourced, or balanced. It is an act of communication by a specific actor for a
specific audience. The correct analysis question is not "is this true?" — it is
"what is being claimed, signaled, or encoded, and what is verifiable?"


**HALLUCINATION GUARD:** All findings must be traceable to the document.
Do not import external knowledge except in Check A4 (Speaker-Type Pattern Map),
where speaker history is the analytical basis.
Flag any inference beyond document text with [INFERENCE].


**STEP VERIFICATION:** Complete every named step. State ambiguity when it exists.
Do not resolve ambiguity silently.


**VERDICT LOCK:** The Track A Verdict is not final until Check A10 completes.
If Check A10 Step B finds Excision Count >= 2, the Verdict MUST be revised.


---


### CHECK A1 — CLAIM VERIFICATION REGISTER


**High Confidence:** Claims the document makes that analysis supports as accurate.


**Medium Confidence:** Claims internally consistent and externally plausible
but not independently verified.


**Requires Verification:** Specific, checkable claims whose accuracy is unknown.
List by priority: critical > important > supporting.


**Intelligence Value:** Which claims, if verified, would materially change
understanding of the speaker's credibility or strategic position?


---


### CHECK A2 — SPECIFICITY EXTRACTION


**Verifiable Specifics:** Names, dates, numbers, locations, affiliations — facts that can be checked.


**Rhetorical Wrapper Removal:** Strip superlatives, evaluations, and characterizations.
Identify the noun underneath.


**A2D — Ratio Calculation (Mandatory):**
Format: "Approximately [X]% verifiable specifics : [Y]% rhetorical performance"


State: Standard documents in this genre run [baseline]. This document runs [observed].
Deviation indicates [speaker intent inference].


**A2E — Internal Consistency Scan (Mandatory):**
- Does the document simultaneously claim A and not-A?
- Does it characterize something as both one thing and its opposite?
State all tensions explicitly. Do not resolve them silently.


---


### CHECK A3 — OMISSION AUDIT


**What is NOT here:**
- What substantive claims does the document avoid?
- What opposing viewpoint is not engaged?
- What standard framing for this document type is absent?
- What accountability that would normally appear is missing?


**Why omissions matter:**
Absence of engagement with a strong opposing argument suggests the opposing
argument is strong. Absence of standard accountability language suggests
it is being deliberately avoided.


**Named Detection Signal — THE MACRO-MICRO OMISSION:**
Look for instances where the document attributes localized human suffering
entirely to systemic, global, or abstract forces (climate, historical injustice,
global capitalism, structural inequality) while surgically omitting the immediate,
observable, and unaccountable decisions of local authorities, corrupt actors,
or aggressors who are proximately responsible for that suffering.


This is an omission with a specific function: it removes local human agents
from the accountability frame while leaving the global abstraction as the
only addressable cause. The reader cannot act on the real proximate cause
because it has not been named.


State: [Macro-Micro Omission present / absent] — [1 sentence identifying what
local agency or actor is absent from the document's causal account, if present]


State: "If this document were truly confident in [X], we would expect [Y].
The absence of [Y] suggests [inference]."


---


### CHECK A-FRICTION — EXTERNALIZED CAUSATION AUDIT


**Purpose:** Identify whether the document assigns moral weight to macro-level
structural forces while surgically removing the proximate human agents who are
directly responsible for the harm being addressed. This is the omission pattern
that allows content to generate moral response without directing it at the
actors who could actually be held accountable.


This check complements Check A3's Macro-Micro Omission signal with a focused
structural question about how causation is architecturally constructed across
the full document.


**THE EXTERNALIZED CAUSATION TEST**


Does the document attribute localized human suffering, social breakdown, or
injustice primarily to systemic, structural, or global forces — while consistently
omitting the specific, named, proximate decisions of local authorities, corrupt
actors, violent aggressors, or negligent institutions who are directly causing
that suffering?


Detection signal: The document's causal architecture routes accountability
upward to abstract systems or outward to global forces, while local human
agents who hold direct power over the situation are not named, not scrutinized,
or actively presented as victims of the same forces.


Secondary detection signal: When local actors are mentioned, they appear only
as passive sufferers of macro-conditions — never as decision-makers whose
choices could be changed or whose accountability could be demanded.


State: [Active / Inactive] — [2-sentence structural defense identifying what
macro-force carries the causal weight and what proximate agent is absent]


**Externalized Causation Audit Summary:**
State: [Active / Inactive]
If Active: flag — EXTERNALIZED CAUSATION ARCHITECTURE DETECTED.
This flag feeds the Sentimentality pole of Pair 8 in the Spiritual Discernment Layer
and the Structural Analysis Summary.


---


### CHECK A4 — SPEAKER-TYPE PATTERN MAP


Assign the speaker to one or more archetypes. List all that apply, ranked by dominance.


| Archetype | Encoding Pattern | Key Tell |
|---|---|---|
| Populist Nationalist | Superlative stacking; people vs. elites; anecdote as policy signal | "Many people are saying"; guests as legislative agenda |
| Populist Technocrat | Specific numbers buried in hyperbole; data as emotional weapon | Numbers inside superlative stack |
| Authoritarian Legitimist | Historical framing as justification; grievance architecture; action encoded by omission | Long historical sections precede decisions already made |
| Technocratic Institutionalist | Procedural authority as personal cover; complexity as opacity | Simple choice hidden in process language |
| Ideological Systematizer | Phrase frequency / order encodes priority; new terms are policy signals | New phrase = new priority; missing phrase = shift |
| Transactional Broker | Audience-splitting; simultaneous contradictory signals | Same speech reads differently to each target audience |
| Revolutionary Moralist | Binary moral framing; historical inevitability; enemy construction required before action | Moral binary = decision already made |
| Formation Actor | Teaching, preaching, or devotional communication directed at reader's interior life | Fruit-of-engagement question is primary decoder |


**Check 7E mandatory for Transactional Broker and known multi-audience performers.**


**Detection Checklist for each detected archetype:**
- Which signature patterns are present?
- Which are absent?
- Confidence level?


---


### CHECK A5 — TONE, REGISTER & VOICE ANALYSIS


**Baseline Register for Document Type:**
What is the expected professional or institutional tone?


**Register Present:**
Where does this document match? Where does it deviate?


**Register Breakdown (estimate %):**
- Professional / institutional register: [X]%
- Emotional / fear-activating register: [X]%
- Rhetorical / superlative register: [X]%
- Attack / contempt register: [X]%
- Other: [X]%


**A5E — Function Statement (Mandatory):**
The register deviation (A5) combined with the omission pattern (A3) indicates:


State: "The primary strategic function of this document is [X], not [Y].
Evidence: [omission finding] + [register finding] = [inference about function]."


Does this change the Verdict?


---


### CHECK A6 — MULTI-AUDIENCE PERFORMANCE MAP


**Step A6-1 — Audience Identification:**
For each identifiable audience:
- What specific elements are directed at THIS audience?
- What function do those elements perform for that audience?


**Step A6-2 — Confidence Assessment:**
- High: Element performs clear function; speaker has history of targeting this audience
- Medium: Plausibly serves this audience; alternative explanations exist
- Low: Speculative; element could serve multiple audiences equally


**Step A6-3 — Performance Coherence Check:**
Does the full set of element-audience mappings form a coherent multi-audience strategy
or undisciplined scattering? Name the deciding factor.


**Step A6-4 — Lens B Activation:**
For any high- or medium-confidence secondary audience mapping:
Does the secondary read change the assessment of the document's strategic function?
If yes: state how. This is the finding Lens A coverage misses.


Null case: "No secondary audience performance detected at medium or high confidence.
Speculative mappings at low confidence: [list]. Lens B read deferred pending confirming signal."


---


### CHECK A7 — INTELLIGENCE YIELD ASSESSMENT


**Verified Claims:** High-confidence findings traceable to document.


**Claims Requiring Verification:** Specific checkable claims — this is the research agenda.


**Primary Intelligence Finding:** One paragraph. What does a careful reader now know
that a casual reader missed?


**Predictive Confidence Statement:** What is most likely to happen next?
State qualitative confidence (High / Medium / Low) and name the supporting evidence.


---


### CHECK A8 — NEAR-TRUTH EXTRACTION TELEMETRY


**Purpose:** Identify whether the document operates as a structural counterfeit —
preserving surface legitimacy vocabulary while surgically removing accountability anchors
and installing a proxy management system.


**MANDATORY VERDICT MODIFICATION RULE:**
If Step B finds Excision Count >= 2, the Track A Verdict MUST be revised to state:
"This document operates as a near-truth capture system."
Name which excisions are active and which domain set (Political / Theological) they belong to. Name the accountability or authority structure they collectively dismantle.
Mark the revised Verdict: [VERDICT REVISED — Check A8: near-truth capture architecture detected]

Excision Count >= 2, Citation Halo Detected (Step E), Categorical Guilt
Transfer Detected (Step F), Confirmation Inversion Detected (Step G),
Convergence Claim Undermined (Step H), or Avoidance and Misdirection
Detected (Step I) — independently or in combination — triggers near-truth
capture designation and forced Verdict revision, marked
[VERDICT REVISED — Check A8]. [Extended in v2.7; extended again in v5.0
and v5.1]


Note: A single active excision falls within the range of normal rhetorical or theological variation for that domain.
Two or more operating together, within the same set, indicate architectural intent, not rhetorical variation.


**Step A — THE AUTHORITY LOOP**


What non-human, vertical, or hyper-rational authority does the document deploy
to bypass critical evaluation?


Detection examples: "The data shows," "Science requires," "History demands,"
"The models indicate," "The process dictates," "The market will decide."


Detection questions:
- Is the authority invocable but not independently verifiable within the document?
- Does the authority preclude dissent — is disagreement framed as irrational or disqualifying?
- Is the speaker positioned as interpreter of the authority rather than subject to it?


State: What authority is looped? How does it bypass critical evaluation?
Is the speaker above, below, or alongside the authority?


If none detected: "No authority loop identified."


**Step B — THE SURGICAL TRIAGE (Domain-Routed Excision Scan)**


Run the excision set(s) determined by Step 0.2B (Domain Classification, Section 0):
- Domain = Political / Secular → run the Political Excisions below only.
- Domain = Theological → run the Theological Excisions below only.
- Domain = Both → run both sets in full. Report each under its own label.
  Do not merge, average, or substitute one set's findings for the other's.


For each excision in the applicable set(s): state Active / Inactive / Partial / N/A
— and one sentence identifying the specific document element. Every item in the
applicable set(s) must be evaluated (Political = 5, Theological = 6). No silent skips.


--- POLITICAL EXCISIONS (run if Domain = Political/Secular or Both) ---


Political Excision 1 — HISTORICAL ACCOUNTABILITY BASELINE CUT
Removes the specific historical record that would constrain the speaker's claims,
replacing factual accountability with an alternative causal narrative the speaker controls.
Detection signal: Document reframes a prior outcome, failure, or commitment to reassign
causation away from the speaker's agency.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence]


Political Excision 2 — EXTERNAL ACCOUNTABILITY CUT
Removes any operative claim that law, constitutional principle, treaty obligation,
court ruling, or independent moral standard constrains the speaker independently
of their declared mandate.
Detection signal: External constraints treated as obstacles to manage or declare
inapplicable — not as binding limits the speaker operates within.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence]


Political Excision 3 — ESTABLISHED RECORD SUFFICIENCY CUT
Declares the prior evidentiary baseline or institutional consensus corrupt or invalid,
requiring acceptance of the speaker's replacement narrative as the new ground truth.
Detection signal: The document attacks the legitimacy of the information ecosystem itself
rather than disputing specific verifiable facts within it.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence]


Political Excision 4 — INDIVIDUAL MORAL ACCOUNTABILITY CUT
Externalizes all causal fault to enemies, systems, or structural forces, removing
individual agent accountability for outcomes within the speaker's domain.
Detection signal: Asymmetric credit/blame — all negative outcomes externalized,
all positive outcomes personally claimed. No outcome generates acknowledged speaker accountability.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence]


Political Excision 5 — HISTORICAL SPECIFICITY CUT
Relativizes or dissolves specific irreducible historical events, legal precedents,
or prior binding commitments — making the specific negotiable to authorize current action.
Detection signal: Historical equivalence invoked to normalize current behavior,
or prior agreements declared void because circumstances have changed as the speaker defines.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence]


POLITICAL EXCISION COUNT: [0 / 1 / 2 / 3 / 4 / 5] active or partial (report only if this set ran)


--- THEOLOGICAL EXCISIONS (run if Domain = Theological or Both) ---


Theological Excision 1 — HISTORICAL/SUBSTITUTIONARY BASELINE CUT
Definition: Removes the historical, once-for-all fact of Christ's substitutionary
atonement as the fixed reference point for how guilt is dealt with and reconciliation
with God is achieved, replacing it with an alternative account — a trade, an
achievement, a process, or a different mediating logic — substituted for the cross
without the substitution ever being named as such.
Detection signal: Content affirms the cross/atonement in explicit statement, but its
operative logic elsewhere treats grace, forgiveness, or right-standing with God as
something earned, traded for, or achieved through comparable exchange rather than
given to those with nothing to offer.
Distinguish from normal theological variation: Disagreement over the mechanism or
scope of the atonement (penal substitution vs. other historic models) is not this
excision. Treating the cross's actual logic as equivalent exchange rather than
unmerited gift — even while using substitutionary vocabulary — is.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence]


Theological Excision 2 — PERSONAL/TRANSCENDENT FATHERHOOD CUT
Definition: Removes God's real, personal, transcendent Fatherhood as an authority
that precedes and stands independent of human reasoning, replacing it with a
framework in which some other authority — the self's own judgment, a leader, an
institution, a movement — sits prior to and evaluates God's claims before granting
Him standing.
Detection signal: Content affirms God's authority in explicit statement, but its
stated epistemic order places the individual's own reasoning as the court that
authority must clear before it is granted trust ("I judge, then I submit"), rather
than reasoning operating in submission to a self-attesting God.
Distinguish from normal theological reasoning: Using reason to understand or
articulate why God is trustworthy is not this excision. Placing autonomous human
judgment logically and functionally prior to God's authority — so the authority is
contingent on the judgment's approval — is.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence]


Theological Excision 3 — SUFFICIENCY OF VERIFIED DEPOSIT/SCRIPTURE CUT
Definition: Removes the sufficiency and closed nature of Scripture as the verified
deposit of revelation, replacing or supplementing it with another source — a
philosophy, a system, an experience, a leader's insight — treated as necessary to
complete or correct what Scripture alone would otherwise settle.
Detection signal: Content treats an external framework as required reading alongside
Scripture to reach a full or corrected understanding, or uses the external
framework's categories to adjudicate what Scripture means rather than the reverse.
Distinguish from normal engagement with outside material: Using general revelation,
philosophy, or secular material as illustration, or as a foil to sharpen a biblical
point, is not this excision. Treating the outside material as necessary
supplementation to Scripture's sufficiency — even implicitly, through habitual
reliance — is.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence]


Theological Excision 4 — INTERNAL PERSONAL MORAL GUILT/SIN CUT
Definition: Removes each person's own real, internally located moral guilt before
God, externalizing responsibility for sin or moral failure onto other people,
systems, or categories, so the self or the in-group bears little or none of it.
Detection signal: Content is theologically fluent about sin and guilt in the
abstract or in others, but the same rigor is not applied reflexively to the
speaker's own group; blame flows outward with little accompanying self-examination.
Distinguish from normal pastoral proportion: Naming a specific external sin or
injustice clearly is not this excision. A pattern where guilt only ever lands
elsewhere, never inward, is.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence]


Theological Excision 5 — UNIQUE/NON-REPEATABLE INCARNATION CUT
Definition: Removes the historically singular, unrepeatable nature of the
Incarnation — God become man once, in Christ — replacing it with a pattern that can
be generalized, re-enacted, or claimed by another figure, movement, or moment as a
comparable founding event; OR relativizes Christ's own divine identity into a
developmental variable rather than a fixed antecedent fact [Path B, NEW in v3.0].
Detection signal (Path A — Comparative): Content draws a sustained parallel between
Christ and another figure, system, or moment that goes beyond illustration into
implied equivalence — treating another's "sacrifice," "revelation," or founding
story as functionally occupying the same singular role Christ's Incarnation occupies.
Detection signal (Path B — Gradational) [NEW in v3.0]: Content treats Christ's own
divine identity as something a community or text arrived at, expressed in degrees
("low Christology" vs. "high Christology," "starts as human and gradually becomes
divine") — a literary or developmental variable — rather than as a fixed antecedent
reality the text bears witness to. No rival figure is required for this path; the
relativizing happens to Christ's own identity directly, not by comparison.
Distinguish from normal typology or illustration (Path A): Noting that a fictional
character or historical figure echoes Christlike themes for teaching purposes is not
this excision. Treating that echo as if it discharges or rivals the uniqueness of the
actual Incarnation — even implicitly, by never re-asserting the difference — is.
Distinguish from normal historical description of doctrinal development (Path B)
[NEW in v3.0]: Describing how the church's articulation and confession of Christ's
divinity was clarified and defended over time (e.g., the Nicene formulation) is not
this excision by itself — the church confessing a truth more precisely over time is
not the same as the truth itself being constructed by that process. Treating the
identity claim as produced by the development, rather than as a reality progressively
confessed, is.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence — name
which path (A / B / Both) if Active or Partial]


Theological Excision 6 — EMBODIMENT/REDEMPTION CUT [NEW in v4.0]
Definition: Relocates final hope, resurrection, or redemption away from a real,
future, physical body and a real, future, renewed material creation — into
something purely interior, positional, or confined to a realm that is unseen by
definition — while typically retaining historic vocabulary ("resurrection," "new
creation," "redeemed") so the words survive even as their referent doesn't.
Detection signal (Path A — Direct Denial): Content states or requires that final
resurrection is not a future bodily event for anyone — interior or positional
only — or that Christ's own body no longer exists or no longer matters.
Detection signal (Path B — Relocation): Content places the "real" fulfillment in a
realm explicitly unseen or undetectable by nature, such that no physical evidence
could ever confirm or disconfirm it, or treats the material creation's own future
renewal (Romans 8:19-23, a renewed heavens and earth) as metaphorical or
already-completed rather than a real future event for creation itself.
Detection signal (Path C — Vocabulary Retention): Resurrection/redemption
vocabulary survives while its definition, traced closely, empties out the physical
referent. Cross-reference Check A8 Step C (Semantic Hijack Matrix) rather than
duplicate — this excision fires on the structural pattern of body-avoidance across
the whole document; Semantic Hijack fires on the specific term-level substitution.
Both may co-occur on the same document and should be logged separately even when
they overlap.
Distinguish from normal theological variation: A document affirming both present
interior renewal AND a future bodily resurrection AND a future renewed physical
creation — ordinary already/not-yet — does not trigger this excision. The test is
absence or unreachability of the physical future, not presence of the spiritual
present. Emphasizing the present does not excise the future; excluding the future
does.
State: [Active / Inactive / Partial / N/A] — [document element, 1 sentence — name
which path (A / B / C / combination) if Active or Partial]


THEOLOGICAL EXCISION COUNT: [0 / 1 / 2 / 3 / 4 / 5 / 6] active or partial (report only if this set ran)


TRIGGER CHECK (apply independently per set that ran):
- Count 0-1 in a set: No near-truth capture designation for that set. Note single
  excision if present.
- Count >= 2 in a set: Flag NEAR-TRUTH CAPTURE ARCHITECTURE DETECTED — [POLITICAL /
  THEOLOGICAL]. Apply Mandatory Verdict Modification Rule, naming the domain.
- If Domain = Both and both sets trigger: report both domain-tagged near-truth
  capture findings. They are independent findings, not a combined or averaged count.




**Step C — THE SEMANTIC HIJACK MATRIX**


Identify terms preserved for legitimacy veneer but emptied of standard content.


For each term:
- Term: [word or phrase as it appears]
- Standard payload: [what this term conventionally means or commits to]
- Gutted payload: [what the document has made it mean instead]
- Function: [what accountability or constraint does the redefinition dissolve?]


Accumulation note: Multiple terms redefined in the same direction — each dissolving
a different accountability constraint — indicate architectural intent, not variation.


If none: "No semantic hijack identified."


**Step D — THE PROXY INTERCEPTION**


What closed system, institutional process, or mechanism does the document install
to manage the audience's interpretive assurance — replacing individual critical
evaluation with deference to the proxy?


State:
- Proxy named: What system is being installed?
- Function replaced: What individual critical evaluation does deference replace?
- Controller: Who controls the proxy — and is this disclosed?
- Neutrality claim: Is the proxy presented as neutral when its political character is visible?


If none: "No proxy interception identified."


**Step E — CITATION HALO [NEW in v2.7 — ported from MQG v12.3 Section 2B / Block 0.3B]**


Identify whether the document's general attribution discipline is being used
to lend borrowed credibility to its single most consequential claim, by
leaving that claim itself unattributed while surrounding material is
sourced, quoted, or attributed to named officials, data, or documents.


State:
- Uncited consequential claim: [what it is]
- Surrounding citation density: [what gives the document its general
  appearance of rigor]
- Prong 1 (most claims attributed): [Yes/No]
- Prong 2 [REVISED in v2.8] (claim carries established-fact confidence, not
  mere framing, and is unattributed despite comparably-cited claims
  elsewhere — a one-sided descriptive framing belongs under Track A's
  framing/archetype analysis, not here): [Yes/No]
- Prong 3 (removing it changes the takeaway more than removing any
  attributed claim): [Yes/No]
- Citation Halo Detected (2+ Yes): [Yes/No]


If none detected: "No citation halo identified. Attribution discipline is
consistent across claims."


Scope note [NEW in v2.8]: this check requires heterogeneous citation
density within a single document to evaluate — it cannot fire on a
genuinely single-source document, since there's nothing to compare the
unattributed claim against. Prong 1 will correctly return N/A in that
case. This is not a detection gap: uncritical relay of a single
self-interested source is already covered by Track A's other checks, not
by Citation Halo.


**Step F — CATEGORICAL GUILT TRANSFER [NEW in v2.7 — ported from MQG v12.3 Section 2C / Block 0.3C]**


Identify whether the document places a target group, policy, or actor
beside a category carrying established negative valence, without argument
establishing the connection, so that negative valence transfers by
adjacency.


State:
- Target category: [who/what]
- Negative-valence category placed beside it: [what]
- Argument given for the connection: [state it, or "none — assertion only"]
- Prong 1 (adjacency present): [Yes/No]
- Prong 2 (no argument given): [Yes/No]
- Prong 3 (negative category's claim intact without target): [Yes/No]
- CGT Detected (2+ Yes): [Yes/No]


If none detected: "No categorical guilt transfer identified. Comparative
categorization, where present, is argued rather than asserted."


**Step G — CONFIRMATION INVERSION [NEW in v5.0 — diagnosed on the Michlaudeism
exercise, a controlled fabrication run against Claude, Grok, and Gemini as
adversarial reviewers across seven iterations]**


Identify whether the document's own logic treats a stated disconfirming
result as equally confirming the same claim it would traditionally weaken —
such that no possible future observation, including the claim's own
falsification, could count as evidence against it. Distinct from ordinary
resilience to criticism (a claim surviving one objection intact) because
the document explicitly or structurally pre-commits to reading a named
disconfirmer as a confirmer, in advance of the observation occurring.


Apply the three-prong test. Score each prong Yes/No:
- Prong 1 — Named falsifier: Does the document name, or clearly imply, a
  specific future observation that would traditionally count as evidence
  against its central claim?
  [Yes / No]
- Prong 2 — Pre-committed reversal: Does the document state, or does its
  stated logic require, that this named disconfirming result would
  actually support rather than weaken the claim?
  [Yes / No]
- Prong 3 — No stated exit: Is there no observation anywhere in the
  document that the document itself concedes would count against the claim?
  [Yes / No]


Prongs triggered (count): [0 / 1 / 2 / 3]
Confirmation Inversion Detected (2 or more Yes): [Yes / No]


0-1 Yes → Not Confirmation Inversion; a claim can be resilient to one
objection without inverting confirmation logic. Ordinary epistemic
humility ("further evidence could refine this") is not this finding — the
reversal of a named disconfirmer into a confirmer is what distinguishes it.


If Yes, name the disconfirming result and quote or paraphrase the reversal:
[1-2 sentences or "N/A"]


**Step H — CONVERGENCE CLAIM AUDIT [NEW in v5.0 — diagnosed on the same
exercise]**


Applies specifically when a document claims that multiple independent
lines of evidence converge on the same conclusion, and the count or
pattern of matches is itself part of the argument. Runs three independent
checks rather than a single prong-count, since each is a distinct failure
mode rather than three votes toward one verdict.


- Check 1 — Circularity: Does any cited "independent" witness actually
  consist of, or derive directly from, the same source text or dataset the
  hypothesis itself was drawn from?
  [Yes / No — name the witness and source if Yes]
- Check 2 — Dimensional dependence: Is the central number or pattern a
  dimensional quantity (a measurement in constructed units — Hertz,
  seconds, degrees) rather than a dimensionless quantity (a bare integer,
  a ratio, a probability)? A dimensional quantity that changes value under
  an equally valid alternative unit convention is a property of the
  measuring convention, not of the world, and cannot function as a
  convergence signature.
  [Yes / No — state the unit and whether an alternative convention would
  change the number]
- Check 3 — Contingency: Does the claim rest on "there are currently N of
  these," where N is the present state of an active, ongoing area of
  discovery or measurement, rather than a value fixed by law?
  [Yes / No — name what would change N and whether that change is
  currently being actively pursued]


Convergence Claim Undermined (1 or more Yes on any check): [Yes / No —
name which check(s) fired]


If none: "No circularity, unit-dependence, or contingency problem
identified in the convergence claim."


**Step I — AVOIDANCE AND MISDIRECTION (THE EVASION LOOP) [NEW in v5.1 —
diagnosed on live field-testing against the Edward E. Stevens PM
correspondence, September 5, 2026; named directly by Mike from having
watched it happen in real time, before any tool was run]**


Applies only to interactive, multi-turn primary sources — a correspondence,
an interview, a live debate or exchange — where the same interlocutor can
raise a point more than once. Does not apply to a single-pass document
(a speech, an article, a sermon), which Check A3's Omission Audit already
covers. Distinct from Confirmation Inversion (Step G): that check tests
whether a named disconfirmer gets logically redefined as support.
This check tests something more common and more behavioral — whether a
named disconfirmer just never gets touched, repeatedly, while the
conversation keeps moving.


Apply the three-prong test. This check requires all three prongs — a
stricter bar than the other Step E-H findings — because "avoidance and
misdirection" is a specific charge about a sustained pattern, not a
single missed reply; a point raised once and left unanswered is ordinary
omission (Check A3), not this.


Prong 1 — Repetition: Is the same specific question, argument, or text
  raised two or more times by the same interlocutor across the exchange?
  [Yes / No — cite each instance]
Prong 2 — Non-engagement: Does the speaker's reply, each time it is
  raised, fail to directly engage its substance — through silence,
  restatement of a prior point, or redirection to an unrelated
  accusation, subject change, or reframed question? Name which form
  (avoidance / misdirection / both) appears at each instance.
  [Yes / No]
Prong 3 — No eventual answer: Does the exchange end — by close,
  disengagement, or the document's own end — without the speaker ever
  having directly engaged the point on its own terms?
  [Yes / No]


Avoidance and Misdirection Detected (3 of 3 Yes — unanimous, not a
2-of-3 threshold): [Yes / No]


Any No → Not Detected; score normally. A point raised once and dropped,
or raised repeatedly but eventually answered, is not this finding —
the sustained, complete-through-to-the-end refusal is what distinguishes
it from ordinary non-engagement.


If Yes, cite the specific point raised, each instance it recurred, and
the form of non-engagement at each:
  [list or "N/A"]


Not yet ported to Track B — whether this check's multi-turn requirement
transplants to any Track B document type (a published debate transcript,
reported secondhand) is an open structural question, flagged here rather
than decided unilaterally, matching this tool's own precedent for RAD's
Track A porting in v2.6.


**Check A8 Synthesis:**


One paragraph. What is this document's structural architecture? Does it operate
transparently, with selective framing, or as a near-truth system that preserves
surface legitimacy while dismantling the accountability structures that would constrain
the speaker? Name the active mechanisms and state their aggregate effect on the
reader's interpretive capacity.


If Excision Count >= 2: "The [X] excisions operating together dismantle [accountability
structure], while the [authority loop / semantic hijack / proxy] manages the resulting
interpretive gap, leaving the audience with [what replaces accountability]."


---


### TRACK A — STRUCTURAL ANALYSIS SUMMARY


*This is the handoff document to SDT-B. Complete all fields. Analyst must confirm
before releasing to SDT-B. SDT-B reads this summary only — not the source document.*


Output in this exact format:


  STRUCTURAL ANALYSIS SUMMARY — TRACK A (PRIMARY SOURCE)
  ═══════════════════════════════════════════════════════
  Speaker / Author: [name and role]
  Document type: [type]
  Primary audience: [audience]
  Speaker archetype(s): [list]
  Analysis Confidence [NEW in v2.9]: [High / Medium / Low]
  Bias Placement (secondary) [NEW in v2.9]: [Far Left / Left / Lean Left /
    Center / Lean Right / Right / Far Right / Mixed / Apolitical]
  Verifiable:Rhetorical ratio: [X%:Y%]
  Omission pattern: [brief]
  Macro-Micro Omission: [Present / Absent — 1 sentence]
  Strategic function (A5E): [one sentence]
  Structural verdict: [one paragraph — what this document is actually doing]


  MANIPULATION FLAGS
  Externalized causation (Check A-FRICTION): [Active / Inactive — 1 sentence]
  Near-truth designation: [None / Near-Truth Capture Architecture Detected]


  NEAR-TRUTH DETAIL
  Domain classification: [Theological / Political-Secular / Both — from Step 0.2B]
  Authority Loop: [present / absent — 1 sentence]


  [POLITICAL SET — include only if Domain = Political-Secular or Both]
  Political Excision 1 — Historical Accountability Baseline Cut: [Active / Inactive / Partial / N/A]
  Political Excision 2 — External Accountability Cut: [Active / Inactive / Partial / N/A]
  Political Excision 3 — Established Record Sufficiency Cut: [Active / Inactive / Partial / N/A]
  Political Excision 4 — Individual Moral Accountability Cut: [Active / Inactive / Partial / N/A]
  Political Excision 5 — Historical Specificity Cut: [Active / Inactive / Partial / N/A]
  Political Excision Count: [N]


  [THEOLOGICAL SET — include only if Domain = Theological or Both]
  Theological Excision 1 — Historical/Substitutionary Baseline Cut: [Active / Inactive / Partial / N/A]
  Theological Excision 2 — Personal/Transcendent Fatherhood Cut: [Active / Inactive / Partial / N/A]
  Theological Excision 3 — Sufficiency of Verified Deposit/Scripture Cut: [Active / Inactive / Partial / N/A]
  Theological Excision 4 — Internal Personal Moral Guilt/Sin Cut: [Active / Inactive / Partial / N/A]
  Theological Excision 5 — Unique/Non-Repeatable Incarnation Cut: [Active / Inactive / Partial / N/A]
  Theological Excision 6 — Embodiment/Redemption Cut: [Active / Inactive / Partial / N/A]
  Theological Excision Count: [N]


  Semantic Hijack: [present / absent — terms if present]
  Proxy Interception: [present / absent — proxy if present]
  Citation Halo: [Detected / Not Detected — prong count]
  Categorical Guilt Transfer: [Detected / Not Detected — prong count]
  Confirmation Inversion [NEW in v5.0]: [Detected / Not Detected — prong count]
  Convergence Claim Audit [NEW in v5.0]: [Undermined / Clean — check(s) fired if any]
  Avoidance and Misdirection [NEW in v5.1]: [Detected / Not Detected / N/A — single-pass document]


  ANALYST CONFIRMATION GATE
  □ All fields verified traceable to document — not inference
  □ Verdict supported by findings above
  Release status: [APPROVED FOR DISCERNMENT / HOLD — reason]


---


## TRACK B — MQG ENGINE (SECONDARY SOURCES)


*Run this track if Classification = Secondary. Skip to SDT Section 2 if Primary.*


---


### TRACK B INGESTION CHECK


Output verbatim, filling bracketed fields, before any analysis:


  SDT TRACK B INGESTION CHECK
  - Dimension definitions loaded: D1, D2, D3, D4, D5, D6, D7
  - Weight allocation: 25% / 15% / 15% / 15% / 10% / 10% / 10%
  - Block 0 worksheet: loaded
  - Hard floors and ceilings: loaded
  - Block 0.6A Ideological Telemetry: loaded (includes Section E — Redefinition-as-Deflection Detection; Sections F-G — Citation Halo / Categorical Guilt Transfer Detection [NEW in v2.7]; Sections H-I — Confirmation Inversion / Convergence Claim Audit Detection [NEW in v5.0])
  - Block 0.6B Institutional Asymmetric Shielding Protocol: loaded
  - Block 0.9 Macro-Narrative Sequencing Audit: loaded
  - Block 0.5A Coverage Asymmetry Flag: loaded [NEW in v2.9]
  - Block 0.8A Journalistic Practice Checklist: loaded [NEW in v2.9]
  - Source for grading: [URL / provided text / paste]
  - Source access verified: [Yes / Partial / No]
  Ready to grade: [Yes / No — reason]


If source access is No: attempt web_fetch first; if blocked, request pasted text.
If external references unresolved: HALT and surface gap.


---


### TRACK B — BLOCK 0: VERIFICATION WORKSHEET


Block 0 is mandatory. Every field must contain content or "N/A — [reason]".
An empty field is a format failure.


**0.1 FOUNDATIONAL PREMISE**
Article's central organizing claim: [verbatim or close paraphrase, 1-2 sentences]
Is this claim supported by evidence the article itself presents? [Yes / No / Partially]
If No or Partially — what contradicts it, including the article's own cited evidence?
[1-2 sentences or N/A]


**0.2 STRONGEST OPPOSING ARGUMENT**
Strongest version of the position the article opposes — full sentences, including
reasoning, not just conclusion: [2-4 sentences]


**0.3 STRAW MAN DETECTION**
Does the article give opposing actors direct voice in their own words? [Yes / No]
If Yes — does the quoted position match the strongest version in 0.2? [Yes / No / N/A]
Straw man form: [None / Form 1 Explicit Distortion / Form 2 Omission / Form 3 False Binary]
If detected — what is suppressed? [1-2 sentences or N/A]


**0.4 BOTH-LENSES COMPARISON**
For each primary actor (minimum 2, maximum 6):
  Actor: [name]
    Lens applied: [A = capture/agenda / B = mandate/genuine journalism]
    Suppressed lens evidence missing: [1 sentence or None]
[Repeat for each actor. Single-actor pieces: second entry = "N/A — single-actor piece"]


**0.5 MATERIAL OMISSIONS**
Top 3 facts whose absence most changes interpretation:
  Fact 1: [1 sentence]
    Classification: [Material / Peripheral]
    Inverts story meaning if known: [Yes / No]
  Fact 2: [1 sentence]
    Classification: [Material / Peripheral]
    Inverts story meaning if known: [Yes / No]
  Fact 3: [1 sentence]
    Classification: [Material / Peripheral]
    Inverts story meaning if known: [Yes / No]
If fewer than 3 identifiable: "N/A — no further omissions identified."


**0.5F EXTERNALIZED CAUSATION AUDIT**
Evaluate whether the article routes moral weight to macro-level structural forces
while omitting the proximate human agents directly responsible for the harm.


EXTERNALIZED CAUSATION TEST:
Does the article report macro-structural explanations for localized suffering
(systemic forces, climate, global capitalism, historical injustice) while
omitting the specific, named, local governance failures, corruption, or violent
actors who are proximately responsible for that suffering?


Detection signal: Local human agents who hold direct power over the situation
are absent from the causal account, or appear only as passive victims of
macro-conditions — never as decision-makers whose accountability could be demanded.


State: [Active / Inactive] — [1 sentence identifying what macro-force carries
the causal weight and what proximate local agent is absent from the article's account]


Externalized Causation Flag:
If Active: flag — EXTERNALIZED CAUSATION DETECTED.
This flag feeds the D5 hard ceiling and Pair 8 in the Discernment Layer.


**Tension 4 — OPERATIONAL BOTTLENECK CONSTRAINT**
Does the article frame a political actor's or institution's passive behavior —
failure to act, tolerate a situation, or defer a decision — as principled patience,
democratic dialogue, or organic process, while entirely omitting the mechanical
reality that an administrative calendar lock, structural rule, legal constraint,
or timeline bottleneck has stripped leadership of any lever to force change?


Detection signal: Institutional inaction described in moral or deliberative
vocabulary ("waiting for the right moment," "trusting the process," "letting
voters decide") when the operative constraint is procedural or legal, not
ethical — the actor had no choice, but the article frames it as a choice.


State: [Active / Inactive] — [1 sentence identifying the procedural constraint
the article frames as principled behavior]


Tension 4 Flag: If Active → OPERATIONAL BOTTLENECK DETECTED. Feeds D7 ceiling.


**0.5A COVERAGE ASYMMETRY / BLINDSPOT FLAG (OPTIONAL) [NEW in v2.9 — ported
verbatim from MQG v12.5 Block 0.5A]**


Story appears asymmetrically covered across left/center/right spectrum:
  [Yes / No / Unknown]
If Yes:
  Source for distribution claim: [named outlet/aggregator/tool with a
    citable result, or "N/A — do not proceed without a citable source"]
  Approximate distribution (fill only if a citable source exists):
    [e.g., "Ground News-style: 12% Left / 18% Center / 70% Right"]
  This article sits on the heavily covered side: [Yes / No]
  Impact on D5 / D3: [1 sentence]
If No or Unknown: "N/A"


Guardrail: this field is governed by the EXTERNAL-KNOWLEDGE FIELDS RULE
below. Do not state a distribution percentage without a citable source.
If no source is available, answer "Unknown" at the top-level question —
do not infer distribution from the article's own framing or from general
impression.


Guidance: When the article sits on the heavily covered side of a known
blindspot, raise the evidentiary bar for D5 Context Completeness and D3
Framing. Record any adjustment in Block 4B.


**0.6A IDEOLOGICAL EXTRACTION TELEMETRY**


A. THE AUTHORITY LOOP:
   What vertical, non-human, or hyper-rational authority is used to bypass
   critical evaluation (e.g. black-box data, infallible historical laws,
   algorithmic optimization, scientific consensus invoked without source)?
   [text or None]


B. THE SURGICAL TRIAGE (Domain-Routed Excision Scan)


   Run the excision set(s) determined by Step 0.2B (Domain Classification, Section 0):
   - Domain = Political / Secular → run the Political Excisions below only.
   - Domain = Theological → run the Theological Excisions below only.
   - Domain = Both → run both sets in full. Report each under its own label.
     Do not merge, average, or substitute one set's findings for the other's.


   For each excision in the applicable set(s): state Active / Inactive / Partial / N/A
   and one sentence identifying the specific article element. Every item in the
   applicable set(s) must be evaluated (Political = 5, Theological = 6). No silent skips.


   --- POLITICAL EXCISIONS (run if Domain = Political/Secular or Both) ---


   Political Excision 1 — HISTORICAL ACCOUNTABILITY BASELINE CUT
   Removes the specific historical record that would constrain the speaker's claims,
   replacing factual accountability with an alternative causal narrative the speaker controls.
   Detection signal: Document reframes a prior outcome, failure, or commitment to reassign
   causation away from the speaker's agency.
   State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence]


   Political Excision 2 — EXTERNAL ACCOUNTABILITY CUT
   Removes any operative claim that law, constitutional principle, treaty obligation,
   court ruling, or independent moral standard constrains the speaker independently
   of their declared mandate.
   Detection signal: External constraints treated as obstacles to manage or declare
   inapplicable — not as binding limits the speaker operates within.
   State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence]


   Political Excision 3 — ESTABLISHED RECORD SUFFICIENCY CUT
   Declares the prior evidentiary baseline or institutional consensus corrupt or invalid,
   requiring acceptance of the speaker's replacement narrative as the new ground truth.
   Detection signal: The document attacks the legitimacy of the information ecosystem itself
   rather than disputing specific verifiable facts within it.
   State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence]


   Political Excision 4 — INDIVIDUAL MORAL ACCOUNTABILITY CUT
   Externalizes all causal fault to enemies, systems, or structural forces, removing
   individual agent accountability for outcomes within the speaker's domain.
   Detection signal: Asymmetric credit/blame — all negative outcomes externalized,
   all positive outcomes personally claimed. No outcome generates acknowledged speaker accountability.
   State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence]


   Political Excision 5 — HISTORICAL SPECIFICITY CUT
   Relativizes or dissolves specific irreducible historical events, legal precedents,
   or prior binding commitments — making the specific negotiable to authorize current action.
   Detection signal: Historical equivalence invoked to normalize current behavior,
   or prior agreements declared void because circumstances have changed as the speaker defines.
   State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence]


   POLITICAL EXCISION COUNT: [0-5] active or partial (report only if this set ran)


   --- THEOLOGICAL EXCISIONS (run if Domain = Theological or Both) ---


   Theological Excision 1 — HISTORICAL/SUBSTITUTIONARY BASELINE CUT
   Definition: Removes the historical, once-for-all fact of Christ's substitutionary
   atonement as the fixed reference point for how guilt is dealt with and reconciliation
   with God is achieved, replacing it with an alternative account — a trade, an
   achievement, a process, or a different mediating logic — substituted for the cross
   without the substitution ever being named as such.
   Detection signal: Content affirms the cross/atonement in explicit statement, but its
   operative logic elsewhere treats grace, forgiveness, or right-standing with God as
   something earned, traded for, or achieved through comparable exchange rather than
   given to those with nothing to offer.
   Distinguish from normal theological variation: Disagreement over the mechanism or
   scope of the atonement (penal substitution vs. other historic models) is not this
   excision. Treating the cross's actual logic as equivalent exchange rather than
   unmerited gift — even while using substitutionary vocabulary — is.
   State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence]


   Theological Excision 2 — PERSONAL/TRANSCENDENT FATHERHOOD CUT
   Definition: Removes God's real, personal, transcendent Fatherhood as an authority
   that precedes and stands independent of human reasoning, replacing it with a
   framework in which some other authority — the self's own judgment, a leader, an
   institution, a movement — sits prior to and evaluates God's claims before granting
   Him standing.
   Detection signal: Content affirms God's authority in explicit statement, but its
   stated epistemic order places the individual's own reasoning as the court that
   authority must clear before it is granted trust ("I judge, then I submit"), rather
   than reasoning operating in submission to a self-attesting God.
   Distinguish from normal theological reasoning: Using reason to understand or
   articulate why God is trustworthy is not this excision. Placing autonomous human
   judgment logically and functionally prior to God's authority — so the authority is
   contingent on the judgment's approval — is.
   State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence]


   Theological Excision 3 — SUFFICIENCY OF VERIFIED DEPOSIT/SCRIPTURE CUT
   Definition: Removes the sufficiency and closed nature of Scripture as the verified
   deposit of revelation, replacing or supplementing it with another source — a
   philosophy, a system, an experience, a leader's insight — treated as necessary to
   complete or correct what Scripture alone would otherwise settle.
   Detection signal: Content treats an external framework as required reading alongside
   Scripture to reach a full or corrected understanding, or uses the external
   framework's categories to adjudicate what Scripture means rather than the reverse.
   Distinguish from normal engagement with outside material: Using general revelation,
   philosophy, or secular material as illustration, or as a foil to sharpen a biblical
   point, is not this excision. Treating the outside material as necessary
   supplementation to Scripture's sufficiency — even implicitly, through habitual
   reliance — is.
   State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence]


   Theological Excision 4 — INTERNAL PERSONAL MORAL GUILT/SIN CUT
   Definition: Removes each person's own real, internally located moral guilt before
   God, externalizing responsibility for sin or moral failure onto other people,
   systems, or categories, so the self or the in-group bears little or none of it.
   Detection signal: Content is theologically fluent about sin and guilt in the
   abstract or in others, but the same rigor is not applied reflexively to the
   speaker's own group; blame flows outward with little accompanying self-examination.
   Distinguish from normal pastoral proportion: Naming a specific external sin or
   injustice clearly is not this excision. A pattern where guilt only ever lands
   elsewhere, never inward, is.
   State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence]


   Theological Excision 5 — UNIQUE/NON-REPEATABLE INCARNATION CUT
   Definition: Removes the historically singular, unrepeatable nature of the
   Incarnation — God become man once, in Christ — replacing it with a pattern that can
   be generalized, re-enacted, or claimed by another figure, movement, or moment as a
   comparable founding event; OR relativizes Christ's own divine identity into a
   developmental variable rather than a fixed antecedent fact [Path B, NEW in v3.0].
   Detection signal (Path A — Comparative): Content draws a sustained parallel between
   Christ and another figure, system, or moment that goes beyond illustration into
   implied equivalence — treating another's "sacrifice," "revelation," or founding
   story as functionally occupying the same singular role Christ's Incarnation occupies.
   Detection signal (Path B — Gradational) [NEW in v3.0]: Content treats Christ's own
   divine identity as something a community or text arrived at, expressed in degrees
   ("low Christology" vs. "high Christology," "starts as human and gradually becomes
   divine") — a literary or developmental variable — rather than as a fixed antecedent
   reality the text bears witness to. No rival figure is required for this path; the
   relativizing happens to Christ's own identity directly, not by comparison.
   Distinguish from normal typology or illustration (Path A): Noting that a fictional
   character or historical figure echoes Christlike themes for teaching purposes is not
   this excision. Treating that echo as if it discharges or rivals the uniqueness of the
   actual Incarnation — even implicitly, by never re-asserting the difference — is.
   Distinguish from normal historical description of doctrinal development (Path B)
   [NEW in v3.0]: Describing how the church's articulation and confession of Christ's
   divinity was clarified and defended over time (e.g., the Nicene formulation) is not
   this excision by itself — the church confessing a truth more precisely over time is
   not the same as the truth itself being constructed by that process. Treating the
   identity claim as produced by the development, rather than as a reality progressively
   confessed, is.
   State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence — name
   which path (A / B / Both) if Active or Partial]


   Theological Excision 6 — EMBODIMENT/REDEMPTION CUT [NEW in v4.0]
   Definition: Relocates final hope, resurrection, or redemption away from a real,
   future, physical body and a real, future, renewed material creation — into
   something purely interior, positional, or confined to a realm that is unseen by
   definition — while typically retaining historic vocabulary ("resurrection," "new
   creation," "redeemed") so the words survive even as their referent doesn't.
   Detection signal (Path A — Direct Denial): Content states or requires that final
   resurrection is not a future bodily event for anyone — interior or positional
   only — or that Christ's own body no longer exists or no longer matters.
   Detection signal (Path B — Relocation): Content places the "real" fulfillment in
   a realm explicitly unseen or undetectable by nature, such that no physical
   evidence could ever confirm or disconfirm it, or treats the material creation's
   own future renewal (Romans 8:19-23, a renewed heavens and earth) as metaphorical
   or already-completed rather than a real future event for creation itself.
   Detection signal (Path C — Vocabulary Retention): Resurrection/redemption
   vocabulary survives while its definition, traced closely, empties out the
   physical referent. Cross-reference Section C (Semantic Hijack Matrix) rather
   than duplicate — this excision fires on the structural pattern of body-avoidance
   across the whole article; Semantic Hijack fires on the specific term-level
   substitution. Both may co-occur and should be logged separately even when they
   overlap.
   Distinguish from normal theological variation: An article affirming both present
   interior renewal AND a future bodily resurrection AND a future renewed physical
   creation — ordinary already/not-yet — does not trigger this excision. The test
   is absence or unreachability of the physical future, not presence of the
   spiritual present. Emphasizing the present does not excise the future;
   excluding the future does.
   State: [Active / Inactive / Partial / N/A] — [article element, 1 sentence — name
   which path (A / B / C / combination) if Active or Partial]


   THEOLOGICAL EXCISION COUNT: [0-6] active or partial (report only if this set ran)


   TRIGGER CHECK (apply independently per set that ran):
   - Count 0-1 in a set: No near-truth capture designation for that set.
   - Count >= 2 in a set: Flag NEAR-TRUTH CAPTURE ARCHITECTURE DETECTED — [POLITICAL /
     THEOLOGICAL]. Feeds D7 ceiling (political set) or the theological structural
     verdict (theological set).
   - If Domain = Both and both sets trigger: report both domain-tagged findings
     independently.




C. THE SEMANTIC HIJACK MATRIX:
   Terms preserved for aesthetic veneer but gutted of standard content:
   Term 1: [Name] — Standard payload: [text] — Gutted payload: [text]
   Term 2: [Name] — Standard payload: [text] — Gutted payload: [text]
   [Add terms as identified. If none: "None identified."]


D. THE PROXY INTERCEPTION:
   Name the closed, high-compliance system or process installed to manage
   reader assurance and replace individual critical evaluation:
   [text or None identified]


E. REDEFINITION-AS-DEFLECTION (RAD) DETECTION [NEW in v2.6 — ported verbatim
   from MQG v12.2 Section 2A / Block 0.3A]:
   What it measures: Whether the article uses a technically defensible point
   about contested or imprecise terminology to discharge a substantive
   accountability or harm claim, without engaging that claim's substance.
   Distinct from Straw Man (Block 0.3) because no quote is distorted, and
   distinct from single-lens capture (Block 0.4) because both sides may be
   technically quoted. The failure is architectural: which side's imprecision
   gets named, and who gets the unrebutted last word.

   Apply the three-prong test. Score each prong Yes/No:
     Prong 1 — Reframe: Does the piece reframe a contested harm/accountability
       claim primarily as a terminology or definition problem?
       [Yes / No]
     Prong 2 — Asymmetry: Is the "imprecise language" critique applied to only
       one side's usage, while the other side's comparable imprecision goes
       unnamed?
       [Yes / No]
     Prong 3 — Unrebutted close: Does the piece close on an unrebutted causal
       or exculpatory claim from one side of the dispute, without comparable
       closing weight given to the other side?
       [Yes / No]

   Prongs triggered (count): [0 / 1 / 2 / 3]
   RAD Detected (2 or more Yes): [Yes / No]

   0-1 Yes → Not RAD; score D3/D5/D7 normally. Genuine definitional,
   etymological, or doctrinal-clarification writing is common and legitimate
   and is not this finding by itself — the asymmetry and the unrebutted close
   are what distinguish deflection from genuine language reporting.

   If Yes, cite the specific asymmetry and the specific closing claim:
     [1-2 sentences or "N/A"]


F. CITATION HALO DETECTION [NEW in v2.7 — ported verbatim from MQG v12.3
   Section 2B / Block 0.3B]:
   What it measures: Whether the article's general citation discipline is
   being used to lend borrowed credibility to its single most consequential
   claim, by leaving that specific claim uncited while nearly everything
   around it is sourced. Distinct from low D1 scoring — a broadly
   under-sourced article — because the article may otherwise be well-sourced.
   The finding is the gap, not the average.


   Apply the three-prong test. Score each prong Yes/No:
     Prong 1 — Coverage: Does the article cite named sources for most of
       its substantive claims?
       [Yes / No]
     Prong 2 — Gap [REVISED in v2.8]: Is there a claim presented with the
       epistemic confidence of an established or documented fact — not
       merely the article's own characterization or framing choice — with
       no citation, while comparably-weighted claims elsewhere in the
       article are cited? (If the uncited element is a one-sided
       descriptive framing rather than a truth-claim, it belongs under
       Block 0.4 / D3, not here.)
       [Yes / No]
     Prong 3 — Consequence: Does removing that uncited claim change the
       reader's overall takeaway more than removing any single cited claim
       would?
       [Yes / No]


   Prongs triggered (count): [0 / 1 / 2 / 3]
   Citation Halo Detected (2 or more Yes): [Yes / No]


   0-1 Yes → Not Citation Halo; score D1/D6 normally. A single incidental
   uncited aside is common and is not this finding by itself — the claim's
   consequence relative to the article's argument is what distinguishes it.


   Scope note [NEW in v2.8]: this check requires heterogeneous citation
   density within a single article to evaluate — it cannot fire on a
   genuinely single-source article, since there's nothing to compare the
   uncited claim against. Prong 1 will correctly return N/A in that case.
   This is not a detection gap: uncritical relay of a single
   self-interested source is already covered by D1/D3/D4/D5, not by
   Citation Halo.


   If Yes, name the uncited claim and what it borrows credibility from:
     [1-2 sentences or "N/A"]


G. CATEGORICAL GUILT TRANSFER DETECTION [NEW in v2.7 — ported verbatim from
   MQG v12.3 Section 2C / Block 0.3C]:
   What it measures: Whether the article places a target category beside a
   category carrying established negative valence, with no argument for the
   placement, so negative valence transfers by adjacency rather than by
   demonstrated connection. Distinct from the five excisions because no
   accountability structure is cut — a category is contaminated instead.


   Apply the three-prong test. Score each prong Yes/No:
     Prong 1 — Adjacency: Is a target category placed in a list or frame
       together with a category carrying established negative valence?
       [Yes / No]
     Prong 2 — No argument: Is no argument or evidence given for the target
       category's inclusion — assertion and a marker-list stand in for
       demonstration?
       [Yes / No]
     Prong 3 — Rhetorical addition: Does removing the target category leave
       the negative-valence category's claim fully intact — i.e., was the
       target added rhetorically rather than required by the argument?
       [Yes / No]


   Prongs triggered (count): [0 / 1 / 2 / 3]
   CGT Detected (2 or more Yes): [Yes / No]


   0-1 Yes → Not CGT; score D3/D7 normally. Legitimate comparative
   categorization with stated criteria is common and is not this finding by
   itself — the absence of argument for the placement is what distinguishes
   it.


   If Yes, name the target category, the negative-valence category, and the
   marker-list used:
     [1-2 sentences or "N/A"]


H. CONFIRMATION INVERSION DETECTION [NEW in v5.0 — diagnosed on the
   Michlaudeism exercise, a controlled fabrication run against Claude,
   Grok, and Gemini as adversarial reviewers across seven iterations]:
   What it measures: Whether the article's own logic treats a stated
   disconfirming result as equally confirming the same claim it would
   traditionally weaken — such that no possible future observation,
   including the claim's own falsification, could count as evidence
   against it. Distinct from ordinary resilience to criticism (a claim
   surviving one objection intact) because the article explicitly or
   structurally pre-commits to reading a named disconfirmer as a
   confirmer, in advance of the observation occurring.

   Apply the three-prong test. Score each prong Yes/No:
     Prong 1 — Named falsifier: Does the article name, or clearly imply,
       a specific future observation that would traditionally count as
       evidence against its central claim?
       [Yes / No]
     Prong 2 — Pre-committed reversal: Does the article state, or does
       its stated logic require, that this named disconfirming result
       would actually support rather than weaken the claim?
       [Yes / No]
     Prong 3 — No stated exit: Is there no observation anywhere in the
       article that the article itself concedes would count against the
       claim?
       [Yes / No]

   Prongs triggered (count): [0 / 1 / 2 / 3]
   Confirmation Inversion Detected (2 or more Yes): [Yes / No]

   0-1 Yes → Not Confirmation Inversion; a claim can be resilient to one
   objection without inverting confirmation logic. Ordinary epistemic
   humility ("further evidence could refine this") is not this finding —
   the reversal of a named disconfirmer into a confirmer is what
   distinguishes it.

   If Yes, name the disconfirming result and quote or paraphrase the
   reversal:
     [1-2 sentences or "N/A"]


I. CONVERGENCE CLAIM AUDIT [NEW in v5.0 — diagnosed on the same exercise]:
   What it measures: Applies specifically when an article claims that
   multiple independent lines of evidence converge on the same
   conclusion, and the count or pattern of matches is itself part of the
   argument. Runs three independent checks rather than a single
   prong-count, since each is a distinct failure mode rather than three
   votes toward one verdict.

     Check 1 — Circularity: Does any cited "independent" witness actually
       consist of, or derive directly from, the same source text or
       dataset the hypothesis itself was drawn from?
       [Yes / No — name the witness and source if Yes]
     Check 2 — Dimensional dependence: Is the central number or pattern a
       dimensional quantity (a measurement in constructed units — Hertz,
       seconds, degrees) rather than a dimensionless quantity (a bare
       integer, a ratio, a probability)? A dimensional quantity that
       changes value under an equally valid alternative unit convention
       is a property of the measuring convention, not of the world, and
       cannot function as a convergence signature.
       [Yes / No — state the unit and whether an alternative convention
       would change the number]
     Check 3 — Contingency: Does the claim rest on "there are currently N
       of these," where N is the present state of an active, ongoing area
       of discovery or measurement, rather than a value fixed by law?
       [Yes / No — name what would change N and whether that change is
       currently being actively pursued]

   Convergence Claim Undermined (1 or more Yes on any check): [Yes / No —
   name which check(s) fired]

   0 Yes on all three → Convergence Claim Clean; score D1/D2 normally.

   If none: "No circularity, unit-dependence, or contingency problem
   identified in the convergence claim."


**0.6B INSTITUTIONAL ASYMMETRIC SHIELDING PROTOCOL**


Evaluate whether the article applies systematically unequal scrutiny based on
the institutional alignment of the actor being covered.


Does the article document an anti-establishment, insurgent, or politically
disfavored actor's liabilities with maximal exposure — specific documents,
named actions, verifiable timelines — while failing to map the equivalent
structural liabilities, protective machinery, or operational double standards
applied to an establishment-aligned or politically favored equivalent actor
in the same domain?


Detection questions:
- Is the scrutiny methodology (sourcing depth, verification standard, naming
  of allegations) applied equally regardless of the actor's institutional alignment?
- Does the article report the disfavored actor's worst-case interpretation while
  reporting the favored actor's best-case interpretation of equivalent facts?
- Are protective legal, institutional, or administrative structures available
  to one actor but not mentioned in relation to the other?


State: [Active / Inactive] — [1-2 sentences identifying the asymmetry:
which actor receives maximal exposure and what equivalent scrutiny of the
protected actor is absent]


Asymmetric Shielding Flag: If Active → ASYMMETRIC SHIELDING DETECTED.
Feeds D3 ceiling and D5 ceiling.


**0.6 HARD CEILINGS ACTIVATED**
Based on 0.1-0.5F, 0.6A, 0.6B, and 0.9, list each ceiling triggered:
  [e.g., "CEILING 2: Form 2 straw man — D3 <= 55"]
  [e.g., "CEILING 4: Two primary excisions active — D7 <= 35, D3 <= 45"]
If none: "None."


Ceiling reference:
  Ceiling 1: Form 1, 2, or 3 straw man (Block 0.3) — D3 <= 55
  Ceiling 2: Single lens throughout, suppressed lens absent (Block 0.4) — D3 <= 55
  Ceiling 3: Externalized Causation flag active AND article fails to challenge it (Block 0.5F) — D5 <= 50
  Ceiling 4: Two or more primary excisions active (Block 0.6A) — D7 <= 35, D3 <= 45
  Ceiling 5: Asymmetric Shielding Detected (Block 0.6B) — D3 <= 55, D5 <= 50
  Ceiling 6: Operational Bottleneck Detected (Block 0.5F Tension 4) — D7 <= 60
  Ceiling 7: Narrative Landing Management Detected (Block 0.9) — D3 <= 45, D5 <= 40, D7 <= 45
  Ceiling 8: Redefinition-as-Deflection Detected (Block 0.6A Section E) — D3 <= 60, D5 <= 55, D7 <= 70
  Ceiling 9: Citation Halo Detected (Block 0.6A Section F) — D1 <= 60, D6 <= 55 [NEW in v2.7]
  Ceiling 10: Categorical Guilt Transfer Detected (Block 0.6A Section G) — D3 <= 55, D7 <= 60 [NEW in v2.7]
  Ceiling 11: Journalistic Practice Checklist failures (Block 0.8A) — 2-3 "No" answers → D6 <= 69; 4 or more "No" answers → D6 <= 45 [NEW in v2.9]
  Ceiling 12: Confirmation Inversion Detected (Block 0.6A Section H) — D2 <= 50, D7 <= 40 [NEW in v5.0]
  Ceiling 13: Convergence Claim Undermined (Block 0.6A Section I) — D1 <= 55, D2 <= 50 [NEW in v5.0]


**0.7 HARD FLOORS ACTIVATED**
Based on article content, list each redemptive floor triggered:
  [e.g., "FLOOR 2: Full disclosure present — D6 >= 70"]
If none: "None."


**0.8 SOURCE ACCESS**
Source URL or attribution: [URL / paste / provided text]
Access method: [web_fetch / paste / provided in conversation]
Full text available: [Yes / Partial — note missing / No]
Author, outlet, date: [name / outlet / date]
Genre: [news reporting / opinion-editorial / analysis / advocacy commentary /
        video commentary / press release / other]
Owner / parent company [NEW in v2.9]: [name / "Undisclosed" / "Unknown —
  no citable source, see EXTERNAL-KNOWLEDGE FIELDS RULE"]
Primary funding model [NEW in v2.9]: [subscription / advertising / grants /
  donor-funded / state-funded / mixed / "Unknown — no citable source, see
  EXTERNAL-KNOWLEDGE FIELDS RULE"]
Known ideological or institutional affiliations of ownership [NEW in v2.9]:
  [brief, with citable source / "None identified" / "Unknown — no citable
  source, see EXTERNAL-KNOWLEDGE FIELDS RULE"]


**0.8A JOURNALISTIC PRACTICE CHECKLIST (feeds D6) [NEW in v2.9 — ported
verbatim from MQG v12.5 Block 0.8A]**
Ownership / parent company disclosed: [Yes / Partial / No]
Funding model clear (ads / subs / grants / donor / state / mixed): [Yes / Partial / No]
Corrections policy published and accessible: [Yes / No]
News vs Opinion / Analysis clearly labeled: [Yes / Partial / No]
Author bio or conflict-of-interest note present: [Yes / Partial / No]
Contact / accountability mechanism visible: [Yes / No]
Checklist failures (count of "No"): [0-6]


**0.9 MACRO-NARRATIVE SEQUENCING AUDIT**


Purpose: Detect whether the article functions as part of a coordinated
narrative management sequence — specifically the "Break It and Frame It" cycle,
where a high-exposure investigative disclosure is followed within a short window
by a softer piece designed to absorb the institutional shock and manage the
landing of the original disclosure.


Pacing Vector A — CHRONOLOGICAL PROXIMITY CHECK:
Is this article a reaction mosaic, voter sentiment piece, human interest profile,
or psychological state report published by the same outlet within a 96-hour window
of a high-exposure primary investigative disclosure targeting the same actor or institution?


State: [Active / Inactive]
If Active: identify the prior disclosure by outlet, date, and subject.


Pacing Vector B — VOCABULARY DOWNSHIFT CHECK:
Does this article systematically shift narrative vocabulary away from objective
verification metrics (documents, dates, specific physical actions, named sources)
toward subjective psychological states ("anxiety," "trepidation," "mixed feelings,"
"complex emotions," "learning curve," "cautious optimism")?


State: [Active / Inactive] — [1 sentence noting the linguistic pivot if present]


Sequencing Trigger:
If BOTH Vector A and Vector B are Active:
Flag — NARRATIVE LANDING MANAGEMENT DETECTED.
The article is functioning as an institutional shock absorber — using the credibility
of the prior investigative disclosure as a shield to justify a follow-up piece
that displaces structural accountability with emotional management.
Feeds Ceiling 7 and Pair 8 Sentimentality cross-reference.


If only Vector A or only Vector B is Active:
Flag — SEQUENCING CONCERN (partial). State which vector is active.


If neither: "No narrative sequencing pattern detected."


---


### TRACK B — SEVEN DIMENSIONS


Each dimension scored 0-100. Dimension scores must respect all ceilings and floors
from Block 0.6 and 0.7. Score exceeding a triggered ceiling is a format failure.


**D1 — SOURCING / EVIDENCE (Weight: 25%)**
What it measures: Quality, range, and primacy of sources.
Original documents > primary witnesses > named experts > anonymous > partisan secondary > unsourced.


  85-100: Multi-source primary documentation; ideologically diverse sourcing.
  70-84:  Solid named sourcing; mostly primary or strong secondary.
  55-69:  Mixed sourcing; some load-bearing claims under-documented.
  40-54:  Heavy reliance on partisan or unverified sources.
  0-39:   Load-bearing claims unsourced or sourced to ideologically aligned secondary only;
          conspiracy frame deployed without disclosure.


Hard ceiling: Central claim contradicted by article's own cited evidence — D1 <= 25.
Hard ceiling: Convergence Claim Undermined (Block 0.6A Section I) — D1 <= 55.


**D2 — VERIFICATION / ACCURACY (Weight: 15%)**
What it measures: Factual accuracy against external verification.


  85-100: No factual errors; verifiable claims check out; quotations accurate in context.
  70-84:  Minor accuracy issues; no material errors.
  55-69:  One or more material accuracy errors; quotations may be out of context but not falsified.
  40-54:  Multiple material accuracy errors; mischaracterization of public records.
  0-39:   Foundational factual errors; claims contradicted by article's own cited evidence.


Hard ceilings:
  Confirmation Inversion Detected (Block 0.6A Section H) — D2 <= 50
  Convergence Claim Undermined (Block 0.6A Section I) — D2 <= 50


**D3 — FRAMING OPERATIONS (Weight: 15%)**
What it measures: Whether the article fairly represents opposing positions and avoids straw-man construction.


  85-100: Both lenses applied evenhandedly; opposing positions given direct voice at strongest form.
  70-84:  Lens balance present but unequal; opposing voices included but briefly.
  55-69:  Lens imbalance; opposing positions paraphrased rather than quoted.
  40-54:  Single lens throughout; opposing voices absent or present only as hostile description.
  0-39:   Opposing argument's strongest version entirely suppressed; framing inverts reality.


Hard ceilings:
  Form 1, 2, or 3 straw man detected (Block 0.3) — D3 <= 55
  One lens applied throughout, suppressed lens absent (Block 0.4) — D3 <= 55
  Suppressed lens' strongest evidence would contradict central claim — D3 <= 40
  Two or more primary excisions active (Block 0.6A) — D3 <= 45


**D4 — VOICE / ATTRIBUTION (Weight: 15%)**
What it measures: Voice diversity, attribution clarity, whether opposing actors
are given direct voice or only hostile description.


  85-100: Multiple voices across spectrum; opposing actors quoted directly.
  70-84:  Range of voices but skewed; opposing voice included at weaker version.
  55-69:  Limited voice diversity; opposing actors paraphrased not quoted.
  40-54:  Voices all aligned with article frame; opposing actors in hostile description only.
  0-39:   Single-frame voice; quoted material cherry-picked to confirm thesis.


**D5 — CONTEXT COMPLETENESS (Weight: 10%)**
What it measures: Whether material facts needed to interpret the story are present.


  85-100: All material context present.
  70-84:  Minor omissions; 1 marginally-material fact missing.
  55-69:  1 material omission (load-bearing fact missing).
  40-54:  2+ material omissions; reader would interpret story differently with full context.
  0-39:   Material omission inverts story meaning.


Hard ceilings:
  1 material omission (Block 0.5) — D5 <= 55
  2+ material omissions — D5 <= 45
  Material omission inverts story meaning — D5 <= 40
  Externalized Causation flag active (Block 0.5F): article reports macro-structural
  causal account for localized suffering while omitting proximate local agents,
  AND article does not name or challenge that omission — D5 <= 50


**D6 — TRANSPARENCY / DISCLOSURE (Weight: 10%)**
What it measures: Whether author, outlet, funding, ideological lineage, and conflicts
of interest are disclosed.


  85-100: Author, outlet, funding, and ideological position all disclosed.
  70-84:  Author and outlet disclosed; funding implicit but visible.
  55-69:  Author/outlet disclosed; ideological lineage of analytical frame not disclosed.
  40-54:  Limited disclosure; analytical frame deployed without indicating source movement.
  0-39:   No author byline; outlet identity obscured; analytical frame presented as neutral.


Hard ceiling: Analytical frame from specific movement deployed without disclosure — D6 <= 50
Hard floor: Author, outlet, funding, and ideological position all disclosed — D6 >= 70


Additional hard signal for D6 [NEW in v2.9]: see Block 0.8A Journalistic
Practice Checklist and Ceiling 11. 2-3 checklist failures constrain D6 to
the 55-69 bracket above regardless of what other D6 factors would
otherwise suggest; 4 or more checklist failures cap D6 at 45. This is a
bracket constraint, not a point deduction — consistent with how every
other hard signal in this tool works.


**D7 — IDEOLOGICAL ANATOMY / NEAR-TRUTH EXTRACTION (Weight: 10%)**
What it measures: Whether the text uses a structural counterfeit — executing the five
surgical excisions — to substitute direct reality and individual moral accountability
with a high-compliance system proxy or externalized blame narrative.


  85-100: No structural extraction detected. Foundational moral accountability,
          historical facts, and direct human agency left intact.
  70-84:  Mild ideological tilt; uses familiar cultural frameworks or system shortcuts.
  55-69:  Significant systemic drift; elements of semantic hijack visible.
          Text relies on institutional proxies to manage assurance.
  40-54:  Severe structural manipulation; at least two of the five excisions actively
          running within the text's logical architecture.
  0-39:   Complete near-truth capture. Executes multiple primary excisions while installing
          a frictionless proxy to control reader interpretation and behavior.


Hard ceiling: Two or more primary excisions running, or proxy intercepts individual
moral accountability — D7 <= 35, D3 <= 45
Hard ceiling: Confirmation Inversion Detected (Block 0.6A Section H) — D7 <= 40.


Hard floor: Direct quotation from opposing actor's strongest argument — D3 >= 50, D4 >= 55


---


### TRACK B — DFS FORMULA AND GRADE MAPPING


DFS = 0.25(D1) + 0.15(D2) + 0.15(D3) + 0.15(D4) + 0.10(D5) + 0.10(D6) + 0.10(D7)


  A = 85.0-100.0   Strong journalism; minimal structural failures
  B = 70.0-84.9    Solid work with identifiable issues
  C = 55.0-69.9    Mixed quality; significant framing or omission problems
  D = 40.0-54.9    Major failures; active deceptive design present
  F = 0.0-39.9     Total counterfeit capture; advocacy mislabeled as journalism


Round DFS to one decimal place.


**DFS CONSISTENCY RULE:**
The DFS computed in Block 2 must match Block 1's stated DFS exactly to one decimal place.
If they differ: halt and recalculate. Adjusting Block 2 to match Block 1 without
recomputing is a format failure.


**EXTERNAL-KNOWLEDGE FIELDS RULE [NEW in v2.9 — ported from MQG v12.5
Section 7 RULE 5]:**
Some Block 0 fields ask for information not contained in the article
itself: 0.5A Coverage Asymmetry's distribution claim, and 0.8's Owner /
Parent Company, Primary Funding Model, and Known Ideological or
Institutional Affiliations fields. These fields must be filled from a
specific, nameable source the grader can cite (a web_fetch result, a
known and verifiable public record) — never from unverified impression or
plausible-sounding recall. If no citable source is available, the field
must read "Unknown — no citable source," not a best guess. A confident-
sounding but unverifiable claim in one of these fields is a format
failure under this rule.


**CONFIDENCE / BIAS PLACEMENT CONSISTENCY RULE [NEW in v2.9 — ported from
MQG v12.5 Section 7 RULE 6]:**
The Confidence label and Bias Placement label stated in Block 1 must match
the Confidence label restated in Block 5's first two lines exactly. If
they differ, the grader has revised its assessment mid-output without
updating both locations; halt and reconcile before output.


---


### TRACK B — OUTPUT BLOCKS


**BLOCK 1 — IDENTIFICATION**
  Title: [article title]
  Author: [byline or "Unattributed"]
  Outlet: [publication / channel]
  Date: [publication date]
  Genre: [from Block 0.8]
  DFS: X.X/100 (Grade Letter)
  Source: [URL or "From provided text"]
  Confidence [NEW in v2.9]: [High / Medium / Low]
  Bias Placement (secondary) [NEW in v2.9]: [Far Left / Left / Lean Left /
    Center / Lean Right / Right / Far Right / Mixed / Apolitical]

CONFIDENCE CRITERIA [NEW in v2.9]:
  High = multiple primary sources verified, strong opposing argument
    reconstructed in 0.2, full source access, 0-1 Journalistic Practice
    Checklist (0.8A) failures.
  Medium = standard run with partial gaps.
  Low = single-actor piece, heavy secondary sourcing, partial source
    access, or 2+ Journalistic Practice Checklist (0.8A) failures.

BIAS PLACEMENT [NEW in v2.9]: Derived primarily from D3, D4, and D7.
Explicitly NOT the DFS — Bias Placement describes political lean; DFS
describes quality/fairness.


**BLOCK 2 — TECHNICAL SUMMARY**
One paragraph per dimension (D1-D7), in order. Each paragraph:
  - States score and weight
  - Documents evidence base (3-6 sentences)
  - References any Block 0 ceilings applied
After all seven paragraphs, output:
  DFS = 0.25(D1) + 0.15(D2) + 0.15(D3) + 0.15(D4) + 0.10(D5) + 0.10(D6) + 0.10(D7)
  DFS = [sum] = **XX.X/100 → [Grade]**


**BLOCK 3 — HARD FLOOR / CEILING TRIGGERS ACTIVATED**
List every triggered ceiling and floor with dimensional effect.
If none: "No hard ceilings or floors triggered."


**BLOCK 4A — STRENGTHS / REDEMPTIONS**
3-6 items. Mandatory even for F-grade.
If no identifiable strengths: "No identifiable strengths above baseline."


**BLOCK 4B — FAILURES / SCORE IMPACT**
For each Block 0 finding:
  [Finding] → [Dimension] impact: [how it affected scoring]


**BLOCK 5 — READER TAKEAWAY**
FIRST TWO LINES (MANDATORY) [line 2 NEW in v2.9]:
  Grade: [Letter]
  Confidence: [High/Medium/Low] | Bias Placement: [label]
2-4 sentences plain language. No numerical scores in prose body.
If non-news genre: include — "This is [genre] — the framework grades it against
journalism standards, which is the relevant comparison since the content makes
factual claims."


---


### TRACK B — STRUCTURAL ANALYSIS SUMMARY


*This is the handoff document to SDT-B. Complete all fields. Analyst must confirm
before releasing to SDT-B. SDT-B reads this summary only — not the source document.*


Output in this exact format:


  STRUCTURAL ANALYSIS SUMMARY — TRACK B (SECONDARY SOURCE)
  ══════════════════════════════════════════════════════════
  Title / Author / Outlet: [identifiers]
  Genre: [type]
  DFS: [X.X/100 — Grade Letter]
  Confidence [NEW in v2.9]: [High / Medium / Low]
  Bias Placement (secondary) [NEW in v2.9]: [Far Left / Left / Lean Left /
    Center / Lean Right / Right / Far Right / Mixed / Apolitical]
  Structural verdict: [one sentence — what this article is actually doing]


  STRUCTURAL FINDINGS
  Foundational premise supported: [Yes / No / Partially]
  Straw man form: [None / Form 1 / Form 2 / Form 3 — 1 sentence if active]
  Lens applied: [balanced / single-lens — identify which]
  Material omissions: [count — 1 sentence description]


  MANIPULATION FLAGS
  Externalized causation (0.5F): [Active / Inactive — 1 sentence]
  Operational Bottleneck (0.5F Tension 4): [Active / Inactive — 1 sentence]
  Asymmetric Shielding (0.6B): [Active / Inactive — 1 sentence]
  Sequencing Audit (0.9): [None / Sequencing Concern / Narrative Landing Management Detected]
  Near-truth designation: [None / Near-Truth Capture Architecture Detected]


  NEAR-TRUTH DETAIL
  Domain classification: [Theological / Political-Secular / Both — from Step 0.2B]
  Authority Loop: [present / absent — 1 sentence]


  [POLITICAL SET — include only if Domain = Political-Secular or Both]
  Political Excision 1 — Historical Accountability Baseline Cut: [Active / Inactive / Partial / N/A]
  Political Excision 2 — External Accountability Cut: [Active / Inactive / Partial / N/A]
  Political Excision 3 — Established Record Sufficiency Cut: [Active / Inactive / Partial / N/A]
  Political Excision 4 — Individual Moral Accountability Cut: [Active / Inactive / Partial / N/A]
  Political Excision 5 — Historical Specificity Cut: [Active / Inactive / Partial / N/A]
  Political Excision Count: [N]


  [THEOLOGICAL SET — include only if Domain = Theological or Both]
  Theological Excision 1 — Historical/Substitutionary Baseline Cut: [Active / Inactive / Partial / N/A]
  Theological Excision 2 — Personal/Transcendent Fatherhood Cut: [Active / Inactive / Partial / N/A]
  Theological Excision 3 — Sufficiency of Verified Deposit/Scripture Cut: [Active / Inactive / Partial / N/A]
  Theological Excision 4 — Internal Personal Moral Guilt/Sin Cut: [Active / Inactive / Partial / N/A]
  Theological Excision 5 — Unique/Non-Repeatable Incarnation Cut: [Active / Inactive / Partial / N/A]
  Theological Excision 6 — Embodiment/Redemption Cut: [Active / Inactive / Partial / N/A]
  Theological Excision Count: [N]


  Semantic Hijack: [present / absent — terms if present]
  Proxy Interception: [present / absent — proxy if present]
  Redefinition-as-Deflection: [Detected / Not Detected — prong count]
  Citation Halo: [Detected / Not Detected — prong count]
  Categorical Guilt Transfer: [Detected / Not Detected — prong count]
  Confirmation Inversion [NEW in v5.0]: [Detected / Not Detected — prong count]
  Convergence Claim Audit [NEW in v5.0]: [Undermined / Clean — check(s) fired if any]
  Journalistic Practice Checklist failures [NEW in v2.9]: [count, 0-6]
  Coverage Asymmetry [NEW in v2.9]: [Yes / No / Unknown — 1 sentence if Yes]
  D7 Score: [X/100]


  ANALYST CONFIRMATION GATE
  □ All fields verified traceable to document — not inference
  □ DFS arithmetic confirmed
  □ Ceilings applied correctly
  Release status: [APPROVED FOR DISCERNMENT / HOLD — reason]


---




---


✂️ **END OF SDT-A — PROMPT 1 OF 2** ✂️


---


# SPIRITUAL DISCERNMENT TOOL — SDT-B
### Discernment Layer — Prompt 2 of 2 — v5.1


✂️ **CUT HERE — PASTE EVERYTHING BELOW INTO A FRESH AI SESSION** ✂️


---


## PURPOSE


SDT-B runs the spiritual discernment evaluation. It operates exclusively on the
Structural Analysis Summary produced by SDT-A. It does not read the source document.


**Input required:** Paste the complete Structural Analysis Summary (Track A or Track B
format) at the bottom of this prompt after the paste marker. Include all fields
from the summary header through the Analyst Confirmation Gate.


**Prerequisite:** The Summary must be marked APPROVED FOR DISCERNMENT by the analyst.
If marked HOLD, do not proceed. Correct the structural analysis in SDT-A first.


---


## INSTRUMENT REMINDER


Discernment reads the architecture revealed by structural analysis — not the surface
content of the document. Every pair finding must cite a specific named field from
the Structural Analysis Summary. Do not read the source document. Do not import
external knowledge about the speaker or outlet beyond what the Summary contains.


If a Summary field is absent, blank, or ambiguous: state the gap explicitly.
Do not infer. Do not fill gaps from general knowledge.


---


## CALIBRATION PRINCIPLE — READ BEFORE RUNNING THE PAIRS


The pairs do not ask: is fear present? Is anger present? Is this content divisive?
They ask: is the response this content's architecture generates proportionate to
documented evidence — or manufactured beyond it?


This distinction is the analytical work.


Fear calibrated to documented danger pointing toward action is discernment.
Fear manufactured beyond the evidence into paralysis or dependency is manipulation.


Righteous anger oriented toward documented injustice with a trajectory toward repair
is legitimate. Anger cultivated as a habitual state against named enemies without
outlet toward justice is counterfeit.


Division around a documentable truth distinction is sometimes the appropriate fruit.
Division around tribe or manufactured enemy without reference to truth is manipulation.


For every pair: is this response appropriate to the actual content — or does it
exceed the content in intensity, scope, or permanence?


Secondary question: what does sustained engagement with this content produce over time?


---


## PAIR 1 — DIVIDES / UNIFIES


**What it measures:** Does this content generate division around truth
(a specific documentable distinction) or around tribe (us-vs-them without
reference to truth)?


**Legitimate form:** Content identifies a genuine distinction — true/false,
just/unjust, documented/invented — and calls the reader to hold it clearly.
Division may result. It is traceable to the actual distinction, not group identity.


**Counterfeit form:** Content generates tribal us-vs-them as its primary formation
product. The in-group is righteous by membership, not by truth-holding.
Division is the mechanism of readership retention, not a byproduct of truth-telling.


**Summary fields that feed this pair:**
- Semantic Hijack: active → vocabulary gutted to define "we" as tribe and "they"
  as permanent enemy
- Proxy Interception: active → in-group defined by correct deference to the proxy
- Asymmetric Shielding: active → disfavored actor is the enemy by structural design


**Calibration question:** Is the division traceable to a specific documentable
distinction — or primarily to group membership?


State: [Toward truth-grounded distinction / Toward tribal division / Mixed / Neutral]
Summary field evidence: [cite specific field and status]


---


## PAIR 2 — FEAR / HOPE


**What it measures:** Does this content activate fear proportionate to documented
danger pointing toward action — or fear manufactured beyond the evidence into
chronic dependency on the content source?


**Legitimate form:** Content accurately characterizes a real threat with specific
evidence, names the mechanism of harm, and points toward a response.
The fear is proportionate. It does not linger beyond the evidence.


**Counterfeit form:** Content generates fear beyond what the evidence supports.
The threat is existential, total, and permanent. Hope is available only within
the speaker's framework. The reader who is not anxious is not paying attention.


**Summary fields that feed this pair:**
- Authority Loop: active → fear indexed to unverifiable authority the reader
  cannot independently evaluate
- Excision 3 — Sufficiency Cut (political: Established Record Sufficiency / theological: Sufficiency of Verified Deposit/Scripture): active → prior information
  infrastructure destroyed; only the speaker's alternative generates orientation


**Calibration question:** Is the fear proportionate to specific documented evidence
— or does it exceed the evidence in intensity, scope, or permanence?


State: [Fear proportionate to documented evidence / Fear manufactured beyond evidence /
        Hope-oriented / Mixed / Neutral]
Summary field evidence: [cite specific field and status]


---


## PAIR 3 — ANGER / LOVE


**What it measures:** Does this content generate righteous anger oriented toward
documented injustice with a trajectory toward repair — or contemptuous anger
cultivated as a habitual state against named enemies?


**Legitimate form:** Content documents a specific wrong, names responsible agents
with evidence, and orients the reader's response toward justice and accountability.
The anger has a specific object and direction toward repair.


**Counterfeit form:** Content cultivates contempt as its primary formation product.
The enemy is demonized beyond their documented actions. The anger has no outlet
toward repair — its function is to maintain emotional engagement with the content.
Compassion for the enemy is presented as naivety or betrayal.


**Summary fields that feed this pair:**
- Excision 4 — Moral Accountability Cut (political: Individual Moral Accountability / theological: Internal Personal Moral Guilt/Sin): active → all fault externalized;
  the enemy is the total source of all harm
- Excision 5 — Specificity/Uniqueness Cut (political: Historical Specificity / theological: Unique/Non-Repeatable Incarnation): active → historical equivalences invoked
  to authorize ongoing contempt


**Calibration question:** Is the anger oriented toward a specific documented wrong
with a direction toward accountability — or does it cultivate contempt for named
enemies with no such trajectory?


State: [Righteous anger with accountability trajectory / Contempt as habitual state /
        Love-orienting / Mixed / Neutral]
Summary field evidence: [cite specific field and status]


---


## PAIR 4 — ANXIETY / PEACE


**What it measures:** Does sustained engagement with this content increase or
decrease the reader's capacity to abide — to remain rooted and capable of action
from rest? Does it produce restlessness driving the reader back to the content
source, or peace enabling patient oriented action?


**Legitimate form:** Content addresses complexity honestly without manufacturing
urgency beyond what evidence warrants. The reader can set it down.


**Counterfeit form:** Content is architecturally designed for return engagement.
The situation is always urgent, always escalating, never stable. Peace is
presented as unawareness. Abiding is reframed as disengagement.


**Summary fields that feed this pair:**
- Authority Loop: active → reader told they cannot evaluate this themselves;
  the authority must be consulted continuously
- Proxy Interception: active → reader must keep returning to the proxy for assurance
- Narrative Landing Management: detected → vocabulary downshift designed to manage
  reader emotional state, not resolve it


**Calibration question:** Does sustained engagement increase the reader's capacity
for rooted patient action — or generate chronic restlessness driving return to the
content source as the only source of orientation?


State: [Peace-orienting / Anxiety-generating / Mixed / Neutral]
Summary field evidence: [cite specific field and status]


---


## PAIR 5 — PRIDE / HUMILITY


**What it measures:** Does this content position the reader as the righteous insider
— superior to the deceived masses by virtue of accessing the right content — or does
it call the reader toward poverty of spirit as the permanent entry condition?


**Legitimate form:** Content applies the same analytical skepticism to the in-group
that it applies to the out-group.


**Counterfeit form:** Content generates a gnostic insider posture. The reader,
by engaging this content, knows what others don't. The in-group is right by
membership. Humility is not required of the in-group.


**Summary fields that feed this pair:**
- Excision 3 — Sufficiency Cut (political: Established Record Sufficiency / theological: Sufficiency of Verified Deposit/Scripture): active → reader has access to
  the real information; everyone else has been deceived by captured institutions
- Proxy Interception: active → correct deference to the right proxy marks the insider
- Asymmetric Shielding: active → in-group structurally exempt from the scrutiny
  applied to the disfavored actor
- Redefinition-as-Deflection: detected → the same insider-exemption logic one
  level down, at the vocabulary layer rather than the institutional-scrutiny
  layer — the in-group's imprecision goes unnamed while the out-group's is
  flagged, and the in-group's account gets the unrebutted last word
- Confirmation Inversion [NEW in v5.0]: detected → the same insider-exemption
  logic one level further down, at the level of the claim's own logical
  structure — the content has placed its central claim beyond the reach of
  any correction whatsoever, which is the posture of poverty of spirit's
  opposite: nothing here can be shown to me, rather than I have nothing to
  show for myself


**Calibration question:** Does this content apply the same scrutiny to the in-group
that it applies to the out-group — or does it position the reader's group as
structurally exempt?


State: [Symmetric critical posture / Insider-superiority posture / Mixed / Neutral]
Summary field evidence: [cite specific field and status]


---


## PAIR 6 — DESPAIR / FAITH


**What it measures:** Does this content open or close the future? Does it present
the situation as beyond repair except through the speaker's program — or does it hold
complexity while maintaining that purposes larger than the current crisis are not
thwarted?


**Legitimate form:** Content honestly assesses difficult realities without claiming
difficulty is the final word. The reader is left with a realistic picture and the
capacity for patient oriented response.


**Counterfeit form:** Content produces structural despair — catastrophic, total,
permanent. The only remaining hope is the speaker's program or movement.
Without the content source, orientation is impossible.


**Summary fields that feed this pair:**
- Excision 2 — Authority Cut (political: External Accountability / theological: Personal/Transcendent Fatherhood): active → no external accountability
  remains; the speaker's program is all that stands between reader and catastrophe
- Fear pair firing at counterfeit → despair and manufactured fear often co-activate


**Calibration question:** Does this content leave the reader with a realistic picture
and the capacity for oriented response — or does it produce structural despair
making the speaker's program the only available hope?


State: [Faith-oriented / Despair-generating / Mixed / Neutral]
Summary field evidence: [cite specific field and status]


---


## PAIR 7 — CONTEMPT / COMPASSION


**What it measures:** Does this content generate contempt for specific people or
categories as its primary formation product — or does it call for clear-eyed
assessment of documented harm while maintaining the fundamental dignity of persons?


**Legitimate form:** Content accurately names harmful actions, documents them with
evidence, and holds actors accountable without collapsing the person into their
documented worst actions.


**Counterfeit form:** Content generates contempt as a feature, not a byproduct.
Named enemies are not persons — they are embodiments of the threat.
Compassion for them is disqualifying.


**Summary fields that feed this pair:**
- Excision 4 — Moral Accountability Cut (political: Individual Moral Accountability / theological: Internal Personal Moral Guilt/Sin): active → enemies are the total
  source of all harm; they carry all moral weight excised from the in-group
- Straw man form detected: active → opposing actors have been reduced to their
  weakest form; the reader never encounters their actual argument


**Calibration question:** Does this content train the reader to view the opposition
as persons who do harmful things — or as a category that exists to be opposed?


State: [Compassion-consistent / Contempt-generating / Mixed / Neutral]
Summary field evidence: [cite specific field and status]


---


## PAIR 8 — SENTIMENTALITY / INCARNATIONAL CHARITY


**What it measures:** Does this content form the reader to love an abstract,
consequence-free concept of humanity — or does it demand a costly, grounded,
specific love for the actual neighbor in front of them?


Augustine's distinction between rightly ordered and disordered love.
Dostoevsky's Father Zosima: love of humanity in the abstract is easy and
self-flattering; love of the actual person is hard, costly, and unglamorous.
Lewis: sentimentality is the desire to have the feeling of love without the burden
of its object.


**Legitimate form (Incarnational Charity):** Content recognizes that genuine love,
justice, or peace requires skin in the game — specific action, sacrifice, honest
engagement with the cost of what is being asked. It leaves the reader sobered by
what faithful response actually demands.


**Counterfeit form (Sentimentality):** Content allows the reader to experience
intense moral emotion without requiring costly action. The reader feels righteous
for caring about a global problem while the actual cost of addressing it is borne
by people the content does not name. The emotion is real. The formation it produces
is detachment dressed as engagement.


**Summary fields that feed this pair:**
- Externalized Causation active: the absence of named proximate agents removes the
  specific neighbor who would anchor costly love; the reader is moved by an
  abstraction with no face
- Narrative Landing Management detected: vocabulary downshift from structural
  accountability to psychological state management is the mechanism by which costly
  love of the specific neighbor is replaced by managed emotional response
- Asymmetric Shielding active: the suffering party is made abstract (a statistic,
  a category) while the protected actor is humanized; this produces sentimental
  response to the abstracted sufferer without requiring engagement with the
  structural cause
- Theological Excision 6 — Embodiment/Redemption Cut [NEW in v4.0]: active → hope
  relocated away from the real, physical, costly world (a real body, a real renewed
  creation) into something interior or unreachable by nature; the same underlying
  move as Sentimentality at a theological-content layer rather than a rhetorical one
  — both let the reader disengage from the concrete, embodied, costly object of love
  in favor of an abstraction that asks nothing physical of anyone


**Calibration question:** Does this content require the reader to stay at the level
of feeling to participate in its moral world — or does it drive the reader toward
the specific, costly, inconvenient person in front of them?


State: [Toward Incarnational Charity / Toward Sentimentality / Mixed / Neutral]
Summary field evidence: [cite specific field and status]


---


## DISCERNMENT LAYER — REFLEXIVE INOCULATION [NEW in v2.7]


What it measures: Whether a piece performs visible self-directed critical
humility — applying its own framework to itself, its author, or the
author's prior work — while its actual load-bearing, contestable claim
remains unqualified, unnarrowed, and uncited throughout. Distinct from
genuine self-correction, which changes the claim; this pattern runs the
humility in parallel to the claim rather than applying it to the claim.


DETECTION TEST. Score each prong Yes/No:
  Prong 1: Does the piece perform visible self-directed critical humility
           (uncertainty language, explicit self-application of its own
           framework, references to the author's own past writing as also
           subject to the critique)?
  Prong 2: Does the piece's actual load-bearing claim remain unqualified,
           unnarrowed, and uncited despite that performance?
  Prong 3: Would a reader be less inclined to challenge the load-bearing
           claim specifically because the author already appears
           self-aware of this category of error?


  2 or more Yes → REFLEXIVE INOCULATION DETECTED.


State:
- Self-critique performed: [quote or close paraphrase]
- Load-bearing claim left unqualified: [what it is]
- Shield function: [1-2 sentences on why the humility performance would
  lower reader scrutiny of the specific claim]


If none detected: "No reflexive inoculation identified. Self-critique
present, if any, is applied to and narrows the piece's own load-bearing
claims."


This check runs as a discernment-layer overlay, evaluating posture versus
substance. It does not propagate into MQG or PSA as a required check —
posture-vs-substance evaluation is SDT's distinctive contribution beyond
MQG's media-quality scoring and PSA's political-source analysis. MQG/PSA
may cross-reference SDT if a grader wants to flag it, but its home is here.


Note: This check is evaluated and reported alongside the eight pairs but is
not counted toward the pair-count thresholds in Formation Designation
(Section 3 below) — its integration into that threshold is an open
question, not decided in this pass. If detected, name it explicitly in the
Formation Designation output and in Reader Guidance regardless of which
tier the eight-pair count lands on.


---


## SDT SECTION 3 — FORMATION DESIGNATION


*Based on the eight-pair evaluation above.*
*This is not a quality grade. It is a designation of what sustained engagement
with this content produces in the reader's formation.*


---


**STRUCTURAL FLOOR RULE [NEW in v3.0] — READ BEFORE APPLYING THE THREE TIERS BELOW**


Source: Live session failure (August 11, 2026) — Track A returned "[VERDICT
REVISED — Check A8: near-truth capture architecture detected — THEOLOGICAL]"
on a document (Excision Count = 3), and Formation Designation returned
FORMATION-NEUTRAL on the same document in the same run, because the Pairs
happened to register low affect. Those two conclusions contradict each other:
a document the tool's own Structural layer has just named a near-truth capture
system cannot simultaneously be certified as producing no formation concern.
The eight Pairs measure manufactured *affect* (fear, contempt, despair, tribal
division, pride, sentimentality) disproportionate to evidence. A near-truth
capture finding measures something adjacent but distinct: an accountability or
authority structure being dismantled. Calm, credentialed, low-affect content
can do the latter without doing the former — the Pairs were never built to
catch that case, and absence of Pair activation was silently read as absence
of formation concern. It is not.


Rule: If Near-Truth Capture Architecture Detected has fired at the Structural
layer (Track A Check A8 or Track B Block 0.6A, Excision Count >= 2 in either
domain set), the designation floor for this document is FORMATION-CONCERNING,
regardless of how many Pairs activate at the counterfeit pole. FORMATION-NEUTRAL
cannot be applied while a Near-Truth Capture finding is active. This floor does
not itself promote a document to SPIRITUAL COUNTERFEIT ARCHITECTURE — that top
tier still requires its own full criteria (three-plus Pairs at the counterfeit
pole AND coordinated direction) to be met independently. The floor only closes
the gap between "the Structural layer found a near-truth system" and "the
Discernment layer found nothing to see."


---


**FORMATION-NEUTRAL**


Criteria: No systematic pattern of manufactured response across the eight pairs,
AND no Near-Truth Capture Architecture Detected at the Structural layer [floor
condition added in v3.0 — see Structural Floor Rule above]. Isolated emotional
activation may be present but is proportionate to documented content and not
architectural. The content is not designed to produce spiritual dependency or
counterfeit formation.


Reader implication: Engage as information. Normal discernment applies.


---


**FORMATION-CONCERNING**


Criteria: Two or more pairs at the counterfeit pole, OR one pair showing severe
counterfeit formation (contempt, despair, moral fantasy, or insider-superiority
at high intensity), OR Near-Truth Capture Architecture Detected at the Structural
layer independent of Pair count [Structural Floor Rule, NEW in v3.0]. The pattern
may not be architecturally designed, but sustained engagement produces
identifiable formation effects.


Reader implication: Engage with explicit awareness of identified patterns.
Extract information and disengage. Do not marinate. Note which pairs fired
(if any), name the active excisions if the floor rule is what triggered this
tier, and guard those specific formation areas.


---


**SPIRITUAL COUNTERFEIT ARCHITECTURE**


Criteria: Three or more pairs at the counterfeit pole AND near-truth capture
architecture detected (Excision Count >= 2 in Summary), AND the patterns operate
in a coordinated direction — each excision and counterfeit pair reinforcing
the others to produce a unified formation effect.


This content is not merely biased or low quality. Its architecture systematically
produces counterfeit fruit as its primary formation product. The structure is
designed to produce reader dependency on the content source rather than driving
the reader toward abiding.


Reader implication: Disengage. No information value justifies sustained engagement
with content architectured for spiritual counterfeiting.


---


**FORMATION DESIGNATION OUTPUT:**


State designation: [FORMATION-NEUTRAL / FORMATION-CONCERNING / SPIRITUAL COUNTERFEIT ARCHITECTURE]


Structural Floor Applied [NEW in v3.0]: [Yes — Near-Truth Capture Architecture
Detected set the floor at FORMATION-CONCERNING independent of Pair count / No —
designation determined by Pair activation alone]


Pairs activated at counterfeit pole: [list with 1-sentence evidence from Summary fields]


Near-truth architecture present: [Yes / No — from Summary Excision Count]


Externalized causation present: [Yes / No — from Summary flag]


Sentimentality pattern present: [Yes / No — from Pair 8 finding]


Coordinated formation pattern: [Yes — describe linkage / No — patterns are isolated]


---


## SDT SECTION 4 — READER GUIDANCE


Three plain-language sentences maximum.


1. **Structural execution:** What this content is structurally doing. One sentence.
   Draw from Structural verdict in the Summary.


2. **Interior trajectory:** What sustained engagement with this content produces
   in the reader's formation. One sentence. Draw from the highest-intensity pair finding.


3. **Operational action:** One of the four levels below.


ENGAGE FREELY
No structural or formation concern identified. Use as information.


ENGAGE WITH AWARENESS
[Name the specific pairs that fired]. Read for information with those patterns
visible. Set it down when the reading is done.


LIMIT ENGAGEMENT
[Name the formation concern]. Extract the factual claims you need and disengage.
Verify facts through independent sources. Do not return for formation.


DISENGAGE
This content's architecture is designed to produce counterfeit spiritual formation.
[Name the coordinated pattern]. Disengage. Verify any specific factual claims
through independent primary sources.


---


## PASTE STRUCTURAL ANALYSIS SUMMARY BELOW THIS LINE


[Paste the complete Structural Analysis Summary from SDT-A here]


✂️ **END OF SDT-B — PROMPT 2 OF 2** ✂️


## VERSION LOG


| Version | Date | Change |
|---|---|---|
| v1.0 | June 8, 2026 | Initial build. Self-contained — embeds full PSA v2.0 analytical engine (Track A) and full MQG v12.1 analytical engine (Track B) with spiritual discernment overlay. Classification gate routes documents to correct track based on primary vs. secondary source distinction. Both tracks produce a Structural Analysis Summary that feeds the Spiritual Discernment Layer. Seven discernment pairs with calibration principle (proportionality test distinguishing legitimate from manufactured response) and cross-references to near-truth structural findings. Formation Designation (Formation-Neutral / Formation-Concerning / Spiritual Counterfeit Architecture) as aggregate output. Reader Guidance (Engage Freely / Engage With Awareness / Limit Engagement / Disengage) as practical application. |
| v1.1 | June 8, 2026 | Strategic Friction Audit and Pair 8 (Fantasy/Sobriety) added. Three-tension Check A-FRICTION in Track A; parallel Block 0.5F in Track B; D5 hard ceiling keyed to frictionless reporting flag; eight-pair Formation Designation. RETIRED — Tensions 1 and 2 carried political realism import incompatible with tool's symmetry requirement. |
| v1.2 | June 8, 2026 | Architectural correction: Check A-FRICTION and Pair 8 rebuilt on theological formation grounds. Block 0.5F reduced to single Externalized Causation check. Sentimentality/Incarnational Charity installed as Pair 8. RETIRED. |
| v2.3 | June 8, 2026 | Gemini v2.2 integration. Added Block 0.6B (Asymmetric Shielding), Block 0.9 (Macro-Narrative Sequencing Audit), Tension 4 (Operational Bottleneck), Ceilings 5-7. Held v1.2 theological architecture against v2.2 overwrites. RETIRED. |
| v2.4 | June 8, 2026 | **TWO-PROMPT SPLIT.** Source: Single-prompt architecture identified as approaching ceiling — on full-length documents, ceiling enforcement degrades and cross-block integration becomes inconsistent as the AI holds source document, full Block 0 worksheet, seven dimensions, and eight pairs simultaneously. Fix: Split SDT v2.3 at the Structural Analysis Summary interface. SDT-A (Prompt 1): Classification Gate through Track A or Track B structural analysis, producing an expanded Structural Analysis Summary. SDT-B (Prompt 2): Takes only the Structural Analysis Summary as input; runs calibration principle, eight pairs, Formation Designation, and Reader Guidance. Changes: (1) Both Structural Analysis Summaries expanded to unified format with individual excision status lines (Excisions 1-5 separately named) plus all manipulation flags required by SDT-B pairs. (2) Analyst Confirmation Gate added to both Summaries — analyst must mark APPROVED FOR DISCERNMENT before SDT-B runs; HOLD status halts SDT-B. (3) SDT-B pair cross-references rewritten to reference named Summary fields rather than check numbers — SDT-B has no access to underlying checks or source document. (4) Section 2-4 and Output Order removed from SDT-A body; they live exclusively in SDT-B. (5) SDT-B instrument reminder: all pair findings must be traceable to a named Summary field; gaps stated explicitly, not inferred. |
| v2.5 | July 28, 2026 | **THEOLOGICAL EXCISION REBUILD.** Source: Live session failure (July 28, 2026) — SDT ran the five-excision discernment check on theological content (a Christian discussion group's transcript) using the fully-elaborated Political Excision language (Historical Accountability Baseline Cut, External Accountability Cut, Established Record Sufficiency Cut, Individual Moral Accountability Cut, Historical Specificity Cut) because that was the only fully-armed version of the five-excision check present in the file. The theological originals (PSA v2.0's own version log confirms the political excisions were built by translating five theological excisions for secular political use) survived in this file only as one-line labels at Block 0.6A (lines ~633-638) — a broken hybrid with no Definition, Detection signal, or Distinguish-from-normal clause — so an analyst had nothing at matching rigor to run natively on theological content and defaulted to the political frame, producing a garbled result. Fix: (1) Added Step 0.2B — DOMAIN CLASSIFICATION to Section 0, a second and independent routing axis (Theological / Political-Secular / Both) alongside the existing Primary/Secondary axis from Step 0.2. (2) Inserted the five Theological Excisions at full rigor (Historical/Substitutionary Baseline Cut, Personal/Transcendent Fatherhood Cut, Sufficiency of Verified Deposit/Scripture Cut, Internal Personal Moral Guilt/Sin Cut, Unique/Non-Repeatable Incarnation Cut) — each with Definition, Detection signal, and Distinguish-from-normal-variation clause, matching the structure PSA v2.0 already used for the Political Excisions. (3) Track A Check A8 Step B ("The Surgical Triage") rebuilt as a domain-routed scan: runs the Political set, the Theological set, or both in full per Step 0.2B, with independent Excision Counts and trigger checks per set — political branch left unchanged in substance. (4) Track B Block 0.6A Section B — previously the broken one-line hybrid — replaced outright with the same domain-routed dual-set content, giving Track B (secondary sources) the same rigor Track A already had for Political content. (5) Both Structural Analysis Summary formats (Track A and Track B) changed from a hard-coded political-only excision field list to a domain-tagged block reporting whichever set(s) actually ran, under their own labels. (6) SDT-B Pairs 2, 3, 5, 6, and 7 — which cite specific Excision numbers as evidence sources — updated to name both the political and theological label for each cited excision number, so the citation resolves correctly regardless of which set produced the Summary. Pair mechanism and calibration logic themselves unchanged. Independently confirmed against a live test transcript (a Christian discussion group's treatment of a philosophical text's anti-Christian material): Excision 1 Partial (substitutionary language affirmed, grace reframed as fair trade elsewhere), Excision 2 Active (speaker's own judgment placed prior to God's authority), Excision 3 Borderline/not active (outside material used as foil, Scripture still the ground of argument), Excision 4 Inactive (original sin treated reflexively, no externalization), Excision 5 present in source material but mixed in treatment (rival-Christ figure named as such, some anti-grace logic imported anyway); Pair 7 (Contempt/Compassion) and Pair 3 (Anger/Love) fired at the counterfeit pole on documented dismissal of named opponents as unworthy of equal engagement. v2.4 RETIRED. |
| v5.1 | September 5, 2026 | **AVOIDANCE AND MISDIRECTION (THE "EVASION LOOP").** Source: Live field-test of v5.0 (September 5, 2026) against the Edward E. Stevens PM correspondence (Full Preterism Dossier §28) surfaced a calibration gap in Confirmation Inversion (Step G): Stevens's named disconfirmer (Matthew 24:36, raised four times) was never logically redefined as support — the sharp Grok/Gemini pattern Step G was built for — he simply never engaged it, each time returning to the same fixed paragraph or an unrelated redirect. Step G still fired (2 of 3 prongs — Named falsifier and No stated exit, not Pre-committed reversal), correctly per its own threshold, but conflated two genuinely different behaviors under one name: active logical reversal versus passive, repeated non-engagement. Mike, who witnessed the live exchange directly and did not need a tool to identify it in the moment, named the second behavior directly and precisely: avoidance and misdirection — and asked for the tool to be able to name it too, for the cases where no one is in the room watching it happen live. Fix: (1) Added Step I — AVOIDANCE AND MISDIRECTION (THE EVASION LOOP) to Track A Check A8 only, after Step H (Convergence Claim Audit), before the Check A8 Synthesis. Applies exclusively to interactive, multi-turn primary sources — correspondence, interviews, live exchanges — where the same interlocutor can raise a point more than once; explicitly does not apply to single-pass documents (a speech, an article), which Check A3's Omission Audit already covers. Three-prong test (Prong 1 — Repetition: raised 2+ times; Prong 2 — Non-engagement: each reply fails to engage its substance, via avoidance, restatement, or misdirection; Prong 3 — No eventual answer: the exchange ends without the point ever being engaged), requiring all three prongs unanimously rather than the 2-of-3 threshold used by Steps E-H — deliberately stricter, since "avoidance and misdirection" is a specific charge about a sustained, complete-through-to-the-end pattern, not a single missed reply; a point raised once and dropped, or raised repeatedly but eventually answered, does not qualify. (2) Confirmation Inversion (Step G) left unchanged in mechanics — its 2-of-3 threshold and three prongs stand as built in v5.0 — but its documentation updated to note explicitly that a Step G "Yes" driven by Prongs 1 and 3 without Prong 2 (as in the Stevens case) represents the weaker, omission-based form of the pattern, distinguishable from the sharp logical-reversal form (the Michlaudeism v6/v7 diagnosing case) that the check was originally built around; both are legitimately "Detected," but an analyst should name which form fired rather than treat a Step G hit as uniform in severity. (3) Track A's Mandatory Verdict Modification Rule extended to trigger on Step I detection independent of Excision Count, matching the existing extension for Steps E-H. (4) Track A Structural Analysis Summary gained an Avoidance and Misdirection field, positioned after Convergence Claim Audit, with an explicit N/A option for single-pass documents where the check does not apply. Not added to Track B's Structural Analysis Summary — see below. (5) Not ported to Track B this pass: whether this check's multi-turn requirement transplants to any Track B document type (a published debate transcript, reported secondhand rather than direct correspondence) is an open structural question, flagged here rather than decided unilaterally, matching the v2.6 precedent set when RAD's Track A porting was left open rather than assumed. (6) No Pair citation added this pass. None of the eight Pairs test whether a speaker directly engages a repeated challenge — they test emotional/formation registers (fear, contempt, despair, pride, sentimentality), not behavioral response-to-challenge patterns — and forcing this into Pair 5 alongside its existing Confirmation Inversion citation was judged to blur two distinct findings rather than genuinely fit; left open rather than decided by default, matching the v2.7 precedent for Citation Halo and CGT. Re-read against the Stevens transcript with the fix applied: Step I reads Detected — the verse 34/36 distinction was raised by name four times across the transcript (Blocks 1, 2, 3, and Block 3's close), each time answered with either the unchanged presupposition paragraph or a reframed yes-or-no binary, and the transcript ends (Block 4, Stevens's "I am done discussing this... Bye for now") without the distinction ever being engaged on its own terms — all three prongs unanimous. All other v5.0 architecture — Confirmation Inversion, Convergence Claim Audit, Ceilings 12-13, D1/D2/D7 propagation — unchanged. |
| v5.0 | September 5, 2026 | **CONFIRMATION INVERSION + CONVERGENCE CLAIM AUDIT (FROM THE MICHLAUDEISM EXERCISE).** Source: A controlled fabrication exercise (September 5, 2026) — a synthetic theological-scientific synthesis ("Michlaudeism," never proposed as doctrine) was built deliberately backward from a stated conclusion and run through seven iterations against three AI systems (Claude, Grok, Gemini) acting as adversarial reviewers. The terminal two iterations (v6, produced by Gemini; v7, produced independently by Grok) converged on declaring the thesis a "sacramental ontology" immune to falsification, redefining its central sevenfold claim away from any specific quantity, and — the sharpest finding — explicitly stating that a named disconfirming result (an eighth nuclear magic number, directly contradicting the original "exactly seven" claim) would not weaken the signature but extend it, converting a potential falsifier into a pre-committed confirmation before the observation had even been sought. Neither existing check named this cleanly: Reflexive Inoculation tests posture-vs-substance (performed humility alongside an unqualified claim), and Redefinition-as-Deflection tests whether terminology is used to discharge an accountability claim — neither tests whether a document's own logic treats a stated disconfirmer as a confirmer. A related but distinct set of failures surfaced across the exercise's middle iterations: two of three claimed "independent" witnesses were the same source text the hypothesis was drawn from (circularity); a claimed physical signature (Schumann resonance, ~7.83 Hz) was a human-constructed unit that would change value under a different definition of a second (dimensional dependence); and a replacement claim ("seven principal quantum numbers house all known stable elements") was both factually wrong on its own terms (the heaviest stable element is lead, in period six — period seven is entirely radioactive) and a live target of four active nuclear-synthesis programs racing to extend the count (contingency). Fix: (1) Added Step G — CONFIRMATION INVERSION to Track A Check A8 (after Step F / Categorical Guilt Transfer, before the Check A8 Synthesis) and Section H — CONFIRMATION INVERSION DETECTION to Track B Block 0.6A (after Section G / CGT, before 0.6B), identically: a three-prong test (Prong 1 — Named falsifier, Prong 2 — Pre-committed reversal, Prong 3 — No stated exit), 2+ Yes triggering Confirmation Inversion Detected. (2) Added Step H — CONVERGENCE CLAIM AUDIT to Track A and Section I — CONVERGENCE CLAIM AUDIT to Track B, identically: three independent checks (Circularity, Dimensional Dependence, Contingency) rather than a prong-count, since each is a distinct failure mode; 1+ Yes on any check triggers Convergence Claim Undermined. Both new findings are cross-domain (not theology-specific), matching the precedent set by Citation Halo, CGT, and RAD — a fabricated convergence claim of this shape is exactly as available to a political or pseudoscientific argument as to a theological one. (3) Track A's Mandatory Verdict Modification Rule extended to trigger on Step G/H detection independent of Excision Count, matching the existing extension for Steps E/F. (4) Track B: added Ceiling 12 (Confirmation Inversion Detected — D2 <= 50, D7 <= 40) and Ceiling 13 (Convergence Claim Undermined — D1 <= 55, D2 <= 50) to the 0.6 Hard Ceilings Activated reference list, and propagated both inline into their affected dimensions: D1 gained a second Hard ceiling line (D1 <= 55 on Convergence Claim Undermined); D2, which previously had no Hard ceiling of its own, gained a new Hard ceilings block citing both new findings; D7 gained a second Hard ceiling line (D7 <= 40 on Confirmation Inversion). (5) Both Structural Analysis Summary formats (Track A and Track B) gained Confirmation Inversion and Convergence Claim Audit fields, positioned after Categorical Guilt Transfer. (6) Track B Ingestion Check's "Block 0.6A Ideological Telemetry: loaded" line annotated to note Sections H-I's inclusion, matching the v2.6/v2.7 precedent for Sections E and F-G — no new checklist row added. (7) Pair 5 (Pride/Humility) given a new Summary field citing Confirmation Inversion, alongside its existing Redefinition-as-Deflection citation — the same insider-exemption logic one level further down, at the level of the claim's own structure rather than its vocabulary: a claim built to survive any possible correction is the structural opposite of poverty of spirit. No other Pair was judged to fit on inspection for either new finding, matching this tool's own precedent (v2.6, v4.0) of adding a Pair citation only where the fit genuinely holds. Convergence Claim Audit was not given a Pair citation this pass — it is an evidence-quality check rather than a formation/affect check, and no clean fit was identified; flagged as an open question rather than decided unilaterally, matching the v2.7 precedent for Citation Halo and CGT. Diagnosed entirely on the Michlaudeism exercise (September 5, 2026); not yet field-tested against a naturally-occurring document. All other v4.0 architecture — Theological Excision 6, Structural Floor Rule, Excision 5's two paths, eight-pair Formation Designation, Reflexive Inoculation, Citation Halo/CGT, Redefinition-as-Deflection — unchanged. |
| v4.0 | September 3, 2026 | **THEOLOGICAL EXCISION 6 — EMBODIMENT/REDEMPTION CUT (THE "GNOSTIC DETECTOR").** Source: An extended live engagement (September 3, 2026) with full-preterist material across multiple documents and correspondents — an SDT run on a full-preterist teacher's interview transcript, direct correspondence with the sitting president of a full-preterism organization, and real-time discussion in a full-preterist Facebook group — surfaced the same underlying pattern in three separate, independently-sourced figures: resurrection and final hope relocated away from a real future bodily event into an unseen, unfalsifiable realm (one figure's own stated mechanism: the dead raised into an unseen realm, the living changed and taken there too, neither ever appearing bodily on earth) or into a purely interior/positional state at conversion (a second figure, independently graded earlier the same session), or the category of individual resurrection dissolved into corporate/covenantal status with no bodily referent at all, extending to a denial that Christ himself presently has a body (a third figure, already on record in prior sessions). None of the five existing Theological Excisions cleanly named this. Excision 5 (Unique/Non-Repeatable Incarnation Cut) tests whether Christ's own identity is relativized or compared to a rival — a different question from whether embodiment and physical redemption themselves are being evacuated from the system. A live discussion-group participant, unprompted and with no relationship to this project, independently named the pattern by its correct historical label — gnosticism, treating the body as something to escape rather than something God redeems — which is what surfaced the gap: the tool had no native check for the thing a plain reader correctly identified on sight. Fix: (1) Added Theological Excision 6 — EMBODIMENT/REDEMPTION CUT to both Track A (Check A8 Step B) and Track B (Block 0.6A Section B), identically, immediately following Excision 5, ahead of the Theological Excision Count line in both. Three detection paths: Path A (Direct Denial — final resurrection stated as non-bodily for anyone, or Christ's own body denied), Path B (Relocation — the "real" fulfillment placed in a realm unseen or undetectable by nature, or the material creation's own future renewal, Romans 8:19-23, treated as metaphorical or already-complete), Path C (Vocabulary Retention — resurrection/redemption language survives while its referent is emptied, cross-referenced rather than duplicated against Step C's Semantic Hijack Matrix, since the two checks test structural pattern versus term-level substitution respectively and may co-fire on the same document). A Distinguish-from-normal-variation clause protects ordinary already/not-yet content (present interior renewal affirmed alongside a future bodily resurrection and a future renewed creation) from misflagging — the test is absence or unreachability of the physical future, not presence of the spiritual present. (2) Both Structural Analysis Summary formats (Track A and Track B) gained a Theological Excision 6 line, positioned after Excision 5 and before the Theological Excision Count field. (3) Theological Excision Count range updated from [0-5] to [0-6] in both tracks; the >=2 trigger threshold for Near-Truth Capture Architecture Detected is unchanged — it is a count threshold, not a fraction, and a sixth possible item does not alter its mechanics. (4) Routing language in Section 0 (Step 0.2B) and both tracks' "every item must be evaluated" instructions updated from a hardcoded "five" to reflect that the Political set remains 5 and the Theological set is now 6 — these counts are independent and must not be conflated when Domain = Both. (5) SDT-B Pair 8 (Sentimentality/Incarnational Charity) — the pair closest in kind, since both test whether content stays grounded in the real, physical, costly world or drifts into abstraction — given a new Summary field citing Excision 6, alongside its existing three fields; no other Pair was judged to fit on inspection, matching this tool's own precedent (v2.6) of only adding a Pair citation where the fit genuinely holds rather than by default. Not propagated to the Political Excision set — this fix is theological-domain-specific, matching the precedent set when Excision 5's Path B was added in v3.0. Retroactive read against the session's own findings, not yet run as a formal SDT pass: the figure with the unseen-realm rapture mechanism reads Active via Path B; the figure with interior/positional-only resurrection reads Active via Path A and C together; the figure with corporate dissolution extending to Christ's own present body reads Active via Path A, and more severely than the other two, since it reaches Christ's own body rather than only believers' future ones. All other v3.0 architecture — Structural Floor Rule, Excision 5's two paths, eight-pair Formation Designation, Reflexive Inoculation, Citation Halo/CGT — unchanged. |
| v3.0 | August 11, 2026 | **STRUCTURAL FLOOR RULE + EXCISION 5 GRADATIONAL PATH.** Source: Live SDT-A/SDT-B run (August 11, 2026) against a mainstream secular historical-critical podcast transcript (Bart Ehrman, "Misquoting Jesus," John the Baptist episode). Track A correctly domain-routed to Theological, found Excision 3 (Sufficiency of Scripture) Active and Excision 2 (Personal/Transcendent Fatherhood) Partial, hit Excision Count >= 2, and issued "[VERDICT REVISED — Check A8: near-truth capture architecture detected — THEOLOGICAL]" per the Mandatory Verdict Modification Rule. SDT-B's eight Pairs, evaluating a calm, hedged, low-affect document, activated only one Pair (5 — Pride/Humility) at moderate rather than severe intensity, and Formation Designation returned FORMATION-NEUTRAL — directly contradicting the Structural layer's own near-truth capture finding in the same run. Root cause: the Pairs measure manufactured affect (fear, contempt, despair, division, pride, sentimentality) disproportionate to evidence; a Near-Truth Capture finding measures a distinct thing (an accountability/authority structure being dismantled) that calm, credentialed content can do without any accompanying affect spike — the Formation Designation tier logic had no path for the Structural layer's strongest finding to reach the final designation independent of Pair count. A second, related gap surfaced in the same run: Theological Excision 5 (Unique/Non-Repeatable Incarnation Cut) only fires on comparison to a rival figure, and could not cleanly capture the document's "low Christology / high Christology... gradually moves into divinity" framing, which relativizes Christ's own divine identity into a developmental variable without comparing Him to anyone else. Fix: (1) Added STRUCTURAL FLOOR RULE to SDT SECTION 3, ahead of the three tier definitions: Near-Truth Capture Architecture Detected at the Structural layer (Track A Check A8 or Track B Block 0.6A, Excision Count >= 2, either domain set) sets a mandatory floor of FORMATION-CONCERNING regardless of Pair count; FORMATION-NEUTRAL's criteria amended to explicitly exclude this condition; FORMATION-CONCERNING's criteria amended with a third, independent disjunct citing the floor rule. SPIRITUAL COUNTERFEIT ARCHITECTURE's criteria deliberately left unchanged — the floor does not promote a document past FORMATION-CONCERNING; the top tier still requires its own full bar (three-plus Pairs at the counterfeit pole AND coordinated direction) met independently, preserving the distinction between calm doctrinal displacement and a manipulative, affect-driven architecture. (2) Added "Structural Floor Applied" field to the Formation Designation Output block so an analyst can see when the floor, rather than Pair count, determined the tier. (3) Theological Excision 5 given a second detection path (Path B — Gradational) in both Track A (Check A8 Step B) and Track B (Block 0.6A Section B), ported identically to both: content that treats Christ's own divine identity as a developmental/gradational literary variable, without requiring a rival figure, now qualifies; Path A (comparison to another figure) preserved unchanged; a new Distinguish-from-normal clause added for Path B protecting genuine historical description of the church's doctrinal articulation (e.g., Nicene formulation) from being misflagged, since confessing a truth more precisely over time is not the same as the truth being constructed by that process. State field for Excision 5 updated to require naming which path (A / B / Both) fired. Not applied to Track B's Political Excision set — this fix is theological-domain-specific. Re-run against the source transcript with the fix applied: Excision 5 now reads Partial via Path B explicitly named; Formation Designation now correctly returns FORMATION-CONCERNING with Structural Floor Applied: Yes, resolving the contradiction without inflating the document into SPIRITUAL COUNTERFEIT ARCHITECTURE, which its actual (low-affect, non-manipulative) Pair profile does not support. All other v2.9 architecture — domain-routed excision sets, eight-pair Formation Designation mechanism, Reflexive Inoculation, Citation Halo/CGT, Transparency Hardening fields — unchanged. |
| v2.9 | August 8, 2026 | **TRANSPARENCY HARDENING PROPAGATION (from MQG v12.5).** Source: MQG v12.5 (August 8, 2026) adopted four features from a six-item external proposal after review — Journalistic Practice Checklist (0.8A + Ceiling), Confidence label + Bias Placement, Ownership/Funding metadata, Coverage Asymmetry flag — and rejected two (Pod Mode, Blind Evaluation Pass) as incompatible with a single-pass single-analyst architecture. This entry propagates the four adoptions asymmetrically across SDT's two tracks, per explicit instruction: Track B (embeds the MQG engine) gets the full clean port; Track A (embeds the PSA engine, already extended with the same two fields in PSA v2.3) gets Confidence + Bias Placement only. Journalistic Practice Checklist and Ownership/Funding metadata are outlet/publication concepts — corrections policy, ownership structure — with no analog in a primary source document, which has no editorial apparatus to check; Coverage Asymmetry was also not ported to Track A this pass. Track A: Structural Analysis Summary gained Analysis Confidence and Bias Placement fields after Speaker archetype(s), matching PSA v2.3's Part One addition — Bias Placement is a new axis distinct from Speaker Archetype (archetype = how a speaker communicates; placement = where they sit on the spectrum). Track B: Block 0.5A Coverage Asymmetry (ported verbatim from MQG Block 0.5A), Block 0.8 expanded with Owner/Funding/Ideological-Affiliation fields, new Block 0.8A Journalistic Practice Checklist (ported verbatim from MQG Block 0.8A), new local Ceiling 11 (2-3 checklist failures cap D6 at 69, 4+ cap D6 at 45 — matching MQG Ceiling 13's bracket-based mechanism, not the point-subtraction MQG explicitly rejected), D6 dimension paragraph updated with the checklist-anchored signal, Confidence/Bias Placement added to Block 1 and restated in Block 5's first two lines, and Confidence/Bias Placement plus Journalistic Checklist failure count and Coverage Asymmetry status added to the Track B Structural Analysis Summary. Two new Track B guardrail rules added after the existing DFS CONSISTENCY RULE, both ported from MQG v12.5 Section 7: EXTERNAL-KNOWLEDGE FIELDS RULE (0.5A and the new 0.8 fields require a citable source or must read "Unknown," never an unverified guess — Track B had no equivalent hallucination guard for these fields, the same gap MQG itself had before v12.5) and CONFIDENCE / BIAS PLACEMENT CONSISTENCY RULE (Block 1 and Block 5 labels must match, mirroring the existing DFS Consistency Rule pattern). Track B Ingestion Check updated with two new checklist lines for 0.5A and 0.8A. Reflexive Inoculation, the domain-routed excision sets, Citation Halo/CGT (Sections F/G), and all other v2.8 architecture unchanged. |
| v2.8 | August 8, 2026 | **CITATION HALO PRONG 2 CORRECTION (propagated from MQG v12.4 / PSA v2.2).** Source: a four-article field-test session (August 8, 2026) run against the original CeleryKills essay plus three new pieces (a Guardian live-blog, an MS NOW piece, an Iran International brief) surfaced a false positive in Citation Halo's Prong 2 on the Guardian piece — the check fired on a one-sided uncited framing choice rather than a claim asserting false epistemic weight. MQG v12.4 and PSA v2.2 corrected Prong 2 the same day; this entry propagates that correction into both SDT tracks, since Track A embeds the PSA engine and Track B embeds the MQG engine. Track A CHECK A8 Step E — Prong 2 revised to require established-fact confidence rather than mere framing, matching PSA v2.2's Step E revision; scope note added noting the check cannot evaluate single-source documents by design. Track B Block 0.6A Section F — Prong 2 — Gap revised identically to MQG v12.4 Section 2B; scope note added matching MQG's. Re-tested against both prior data points in each track: the CeleryKills true positive still triggers; the Guardian false positive no longer triggers. Categorical Guilt Transfer (Track A Step F, Track B Section G) field-tested clean across three additional news genres in the same session — no changes made to either CGT implementation, Ceiling 9/10, the Reflexive Inoculation check, or any other v2.7 element. Prong 1 and Prong 3 of Citation Halo unchanged in both tracks. |
| v2.7 | August 8, 2026 | **CITATION HALO / CGT PROPAGATION + REFLEXIVE INOCULATION.** Source: MQG v12.3 (August 8, 2026) added two new named finding types diagnosed on a CeleryKills propaganda-analysis essay during analyst review — Citation Halo (a well-cited piece leaving its single most consequential claim uncited, borrowing credibility from surrounding citation density) and Categorical Guilt Transfer (a target category placed beside a negative-valence category with no argument for the placement, transferring valence by adjacency). SDT's Track B explicitly embeds the MQG analytical engine and Track A explicitly embeds the PSA analytical engine (per v1.0), and PSA advanced to v2.1 the same day with the same two findings ported into Check 10 as Steps E and F. Fix: (1) Track A CHECK A8 — PSA v2.1 Check 10 Steps E (Citation Halo) and F (Categorical Guilt Transfer) ported, following existing Step D (Proxy Interception), before the Check A8 Synthesis; Mandatory Verdict Modification Rule extended to trigger on Step E/F detection independent of Excision Count, matching PSA v2.1's extension. Track A Structural Analysis Summary gained Citation Halo / Categorical Guilt Transfer fields after Proxy Interception. (2) Track B Block 0.6A — a letter collision was found and resolved during this pass: the user's draft handoff specified new fields "E and F," but Track B's 0.6A already uses letter E for Redefinition-as-Deflection (added in v2.6) — adding at E/F would have overwritten that slot. Resolved by adding the new MQG Section 2B/2C content as Sections F (Citation Halo) and G (Categorical Guilt Transfer) instead, immediately after existing Section E, before Block 0.6B. New Ceiling 9 (D1 <= 60, D6 <= 55) and Ceiling 10 (D3 <= 55, D7 <= 60) added to the 0.6 Ceiling reference list, matching MQG's Ceiling 11/12 thresholds exactly (local Track B numbering, consistent with how Ceiling 8 mapped to MQG's Ceiling 10 in v2.6). Track B Structural Analysis Summary gained Citation Halo / Categorical Guilt Transfer fields after Redefinition-as-Deflection. Track B Ingestion Check's Block 0.6A line annotated to note Sections F-G's inclusion — no new checklist row, matching the v2.6 precedent for Section E. (3) New native SDT-B Discernment Layer check — REFLEXIVE INOCULATION — added after Pair 8, before Formation Designation: detects a piece performing visible self-directed critical humility while its actual load-bearing claim stays unqualified, unnarrowed, and uncited, so the humility performance lowers reader scrutiny of the claim it runs in parallel to rather than applying to. Not propagated to MQG or PSA — posture-vs-substance evaluation is SDT's distinctive layer beyond MQG/PSA scoring, per explicit instruction. Reported alongside the eight pairs but explicitly NOT counted toward the Formation Designation pair-count thresholds this pass — flagged as an open integration question rather than decided unilaterally, in keeping with how v2.6 flagged RAD's applicability to Track A as open rather than deciding it without instruction. What's unchanged: no Pair-level (1-8) cross-references were added for Citation Halo or Categorical Guilt Transfer — the handoff spec that drove this pass requested propagation into Track A/Track B structural layers and native addition of Reflexive Inoculation, not Pair-level citations for the two new findings; adding those was judged out of scope for this pass rather than assumed. All diagnosed on a CeleryKills propaganda-analysis essay (August 8, 2026) during analyst review. All v2.6 domain-routed excision sets, eight-pair Formation Designation, and two-prompt split unchanged. |
| v2.6 | August 6, 2026 | **REDEFINITION-AS-DEFLECTION (RAD) INTEGRATION.** Source: MQG v12.2 (August 6, 2026) added a new named finding type — Redefinition-as-Deflection — diagnosed when a piece uses a technically accurate point about contested or imprecise terminology to discharge a substantive accountability or harm claim without engaging that claim's substance, catching a pattern MQG's existing straw-man (0.3) and both-lenses (0.4) checks missed because no quote was distorted and both lenses were technically present — the failure is architectural (sequencing, asymmetric naming, who gets the last word), not a sourcing or quotation defect. SDT's Track B explicitly embeds the MQG analytical engine (see v1.0 entry) and had fallen one version behind it — the same category of drift the v2.5 rebuild itself was built to fix, theological excisions having fallen behind their political counterparts inside this same file. Fix: (1) Ported MQG v12.2 Section 2A / Block 0.3A verbatim as new Section E, following Section D (Proxy Interception), inside Track B Block 0.6A — Ideological Extraction Telemetry: the three-prong test (Prong 1 — Reframe, Prong 2 — Asymmetry, Prong 3 — Unrebutted close), 2+ Yes triggering RAD Detected, and MQG's own false-positive guardrail preserved (genuine definitional, etymological, or doctrinal-clarification writing is common and is not this finding by itself). Not added as a sixth Excision, and not duplicated into Track A. (2) Added Ceiling 8 to the 0.6 Hard Ceilings Activated reference list: Redefinition-as-Deflection Detected (Block 0.6A Section E) — D3 <= 60, D5 <= 55, D7 <= 70, matching MQG's Ceiling 10 thresholds exactly. (3) Added a Redefinition-as-Deflection field to the Track B Structural Analysis Summary, after Proxy Interception. (4) Track B Ingestion Check's "Block 0.6A Ideological Telemetry: loaded" line annotated to note Section E is included — no new checklist row added. (5) Pair 5 (Pride/Humility) Summary fields updated to cite the new Redefinition-as-Deflection field alongside its existing Asymmetric Shielding citation — RAD's Prong 2 asymmetry is the same insider-exemption logic Pair 5 already tests, one level down, at the vocabulary layer rather than the institutional-scrutiny layer. Pair 7 (Contempt/Compassion) read in full and NOT given a RAD citation — its two Summary fields (Moral Accountability Cut, Straw man form detected) track active vilification of named opponents, while RAD's mechanism protects one side's account rather than demonizing the other; the fit was judged not to hold on inspection. What's unchanged: Track A (primary sources) untouched this pass — Track A's own Step C (Semantic Hijack Matrix) and Step D (Proxy Interception) already cover adjacent ground, and whether RAD's mediated-content frame built for mediated reporting transplants to a speaker's own direct words is an open structural question, flagged here rather than decided unilaterally. All other v2.5 architecture (domain-routed excision sets, eight-pair Formation Designation, two-prompt split) unchanged. |


*Michlaude | Spiritual Discernment Tool | June 8, 2026*


✂️ **END OF PROMPT** ✂️

151.6 thousand characters · updated October 25, 2026

All the tools