Order of Operations — with a Real Example

The following sequence was developed and tested in practice. The example is taken from an actual session evaluating a YouTube channel producing devotional biblical content.

Step 1: Establish the governing environment. Load the Christianize Prompt and the Hallucination Guard at the start of the session. This takes two paste operations and thirty seconds. Every output from this point forward operates from a governed center.

Step 2: Get the transcript. YouTube auto-generates captions on almost all content. Pull it. Thirty seconds.

Step 3: Run the SDT. Is this content worth engaging? What is it forming? The SDT returned a verdict on the channel: theologically orthodox, communicatively strong, worth engaging — but with specific patterns to watch for.

Step 4: Run the Theology Grader. A walkthrough of John 3 from the channel was submitted. Score: 11/30 on theosis depth, Tier C. The architectural diagnosis: the teacher was reading John 3 — a participatory text about origin, nature, and the Spirit's sovereign movement — through a forensic primary lens. The content was accurate. The depth was missing because the architectural lens suppressed it.

For comparison, the same grader was run on the channel's Psalm 91 video. Score: 27/30 — matching the source text point for point. Same teacher. Same channel. Sixteen-point theosis gap between the two videos. That gap is the diagnostic value of the tool. The grader found with precision that the teacher achieves source-text depth when his natural devotional register matches the text's architecture, and loses fifteen points when it does not. The diagnostic is not arbitrary.

Step 5: Ask for specific improvements. After the 11/30 score, the AI was asked: "What specific changes would raise the theosis depth while preserving this teacher's voice and his strongest illustrative material?" The result was a complete revised script — retaining the bronze serpent typology, the Nicodemus three-appearance arc, and the light/darkness section, while correcting the architectural lens and strengthening the participatory frame. A phrase from John 3:21 that the original had passed over — "his deeds have been done in God" — was developed into its own paragraph because the grader had identified it as the chapter's strongest coram Deo moment.

Step 6: Re-score. The revised script scored 25/30. Fourteen-point theosis improvement in a single session.

Step 7: Run the Skeptic. The Skeptic found something the Theology Grader had not caught — an internal logical tension between the sermon's sovereignty argument (new birth is entirely passive, the wind blows where it wills) and its closing invitation (just look, just turn your eyes). If new birth is entirely prior to any capacity in the recipient, the person being invited has no capacity to execute the invitation. The Skeptic named it precisely. One targeted fix resolved it: "this is what the cross declares to those who receive it" — naming the faith posture of reception rather than pretending to an evidential standard the claim was never obligated to meet.

The Skeptic then reached its ceiling. Everything beyond that point was a surveyor's instruments at the edge of the water — calibrated to measure depth, useless for measuring what is in the water or where it comes from. The Skeptic's tools are built for evidence. The claim that new birth is a sovereign act of the Spirit is not an evidence claim. When the instruments have nothing left to measure, the content is standing on ground the instruments were never designed to assess. That is the expected outcome for content that has been well-constructed. When the tool reaches the ceiling, the content is ready.

A second example: the CCB in practice. The example above evaluates content that already exists. The Canonical Context Brief runs earlier still — before the sermon is anything but an assigned text and an open question. The following is the pattern in practice.

Step 1: The pastor chooses his pairing. A pastor preparing to preach finds himself drawn to Numbers 16 and Psalm 84 — not because a message has already been decided, but because he hears the two passages speaking to each other across the Testament line. That choice is pastoral discernment. Nothing in this pipeline automates it, and nothing should.

Step 2: Mode is validated, not assumed. Before the CCB runs a single pass, it asks: "Do you already have a message, or are you letting the text tell you where to go?" The pastor is letting the text lead. No thesis is locked. The brief records this and proceeds in text-first mode — this is expository preaching's native mode.

Step 3: The seven exegetical passes run against both texts. Superscription and authorship, literary context, canonical cross-references, redemptive-historical placement, candidate theological syntheses, Christological trajectory, and the text's own natural address on the disciple's walk — each pass runs against Numbers 16 and against Psalm 84, asking the text-first governing question throughout: what does this text open?

Step 4: The brief returns paths, not a verdict. Two to four candidate directions, each naming its theological architecture, its known seam in the Theology Grader's own vocabulary, and its abiding target. The brief closes with one line and stops: "Which of these, if any, is where you want to go — or does the text pull you somewhere none of these name?" It does not answer its own question.

Step 5: The draft is built already oriented. Whatever the pastor selects, the sermon he eventually writes and submits to the Theology Grader already knows what he knew when he opened to Numbers 16 before he ever preached Psalm 84 — "not because an instrument told him to, but because that is what responsible exegesis has always required." The Theology Grader, the Hallucination Guard, and the Writing Standards pass downstream are unchanged by any of this. The CCB's only claim on the rest of the pipeline is that the draft entering those tools should already be aimed at a known target, rather than discovering one after the fact.