AUDIT_SELECTION_MATRIX.md

# APODICTIC Audit Selection Matrix

<!-- inventory-synced: audits=45:ce0a9a58 research=6:a1b9dfc4 -->

This guide helps route a manuscript to the right APODICTIC passes and specialized audits.

---

## 1) Fast Entry Router

Use this in intake order.

### Step A: What do you have?

1. Idea / notes only -> Use `apodictic-pre-writing` (Pre-Writing Pathway)
2. Partial draft that feels structurally stuck -> Start with **Plot Selection & Coaching** (triage mode)
3. Complete draft or nonfiction material that may not have a story spine -> Continue to Step B (Franklin Pathway may activate)
4. Partial or complete manuscript -> Continue to Step B

### Step B: What do you need most right now?

1. Structural diagnosis and revision priorities -> Run **Core DE**
2. Deep architecture across pacing, POV, theme, continuity -> Run **Full DE**
3. Focused diagnosis on one concern -> Run **Specialized Audit(s)**
4. Market readiness and candid quality verdict -> Add **Pass 11**

### Step C: Which symptom is loudest?

1. "I don't buy character choices" -> **Decision Pressure**
2. "Stakes feel flat" -> **Stakes System**
3. "Scenes feel inert / nothing happens" -> **Scene Turn**
4. "Emotion isn't landing" -> **Emotional Craft**
5. "Action/fights are confusing or empty" -> **Force Architecture**
6. "Reads like AI" -> **AI-Prose Calibration**
7. "Mystery/thriller puzzle isn't working" -> **Mystery/Thriller Architecture**
8. "Worldbuilding is big but not integrated" -> **SFF Worldbuilding Integration**
9. "Characters feel generic or flat" -> **Character Architecture**
10. "Literary prose feels decorative, not structural" -> **Literary Craft**
11. "Series entries don't stand alone" -> **Series & Composite Novel**
12. "Memoir/nonfiction has no narrative shape" -> **Franklin Pathway** (pre-spine viability gate)
13. "I'm stuck / my draft isn't working" -> **Plot Architecture** or **Plot Selection & Coaching** (structural triage)
14. "Need submission-readiness honesty" -> **Pass 11**

**Note on automatic activation:** You don't always need to select audits manually. During a Core or Full DE run, audits activate two ways: (1) **at contract** — genre/mode signals recommend specific audits at intake, and (2) **during passes** — finding-driven triggers recommend audits when pass results indicate the need. See `audit-routing-table.md` for the full activation table.

---

## 2) Core DE Passes (Default Backbone)

Run these for most manuscripts.

| Module | What it does | Use it when |
|---|---|---|
| Pass 0: Reverse Outline | Maps what is literally on the page scene by scene | Always first for objective baseline |
| Pass 1: Reader Experience | Tracks confusion, boredom, delight, and belief breaks | Always; identifies reader-side failure points |
| Pass 2: Structural Mapping | Maps beats, causality, and proportional shape | Always; tests architecture and chain logic |
| Pass 5: Character Audit | Audits wants, fear, agency, voice, and arc coherence | Always; character movement needs verification |
| Pass 8: Reveal Economy | Audits information flow, fairness, and timing | Always; reveal and knowledge logic matter in all modes |

---

## 3) Full DE Add-On Passes

Add when complexity is high or Core DE shows systemic issues.

| Module | What it does | Use it when |
|---|---|---|
| Pass 3: Rhythm and Modulation | Quantifies pacing/rhythm where Pass 1 flagged issues | Pacing stalls or intensity flattening persist |
| Pass 4: Emotional Value Tracking | Tracks valence, intensity, certainty movement | Emotional trajectory feels static or chaotic |
| Pass 6: Scene Function Audit | Tests whether scenes earn space and perform functions | Too many scenes feel redundant or low-impact |
| Pass 7: POV and Voice | Audits POV distribution, distance, and voice consistency | POV drift, head-hop risk, or voice blur appears |
| Pass 9: Thematic Coherence | Tests motif and controlling-idea alignment | Theme drift, thesis pressure, or weak payoff |
| Pass 10: Entity Tracking | Tracks continuity of people, places, rules, facts | Continuity, timeline, world-rule, or state errors |

---

## 4) Pass 11 (Critical Quality + Market Viability)

Use when the author wants publication-readiness or hard verdicts.

| Sub-Pass | What it does | Use it when |
|---|---|---|
| 11A Writing Quality Diagnostic | Prose quality tiering and line-level risk mapping | Any Pass 11 run |
| 11B Critical Verdict Protocol | Multi-lens candid evaluation | Any Pass 11 run |
| 11C Market Reality Check | Market fit, comp quality, acquisition realism | Marketability or category fit is requested |
| 11D First-50 Conversion Gate | Opening-pages conversion and retention pressure test | Query/submission prep |
| 11E Revision Economics | Revision ROI and sequencing strategy | Planning major revision rounds |
| 11F Adversarial Reader Stress Test | Hostile-reader attack surface analysis | Optional hard-truth stress test |

---

## 5) Specialized Audits (Craft and Universal)

Use these when one craft system needs deeper diagnosis than core passes provide.

| Audit | What it does | Use it when |
|---|---|---|
| Stakes System | Tests escalation and consequence architecture | Reader says stakes feel low or repetitive |
| Decision Pressure | Tests choice plausibility under pressure | Reader says character decisions feel unearned |
| Scene Turn | Tests scene-sequel mechanics and turn integrity | "Nothing happens" or scene drag complaints |
| Emotional Craft | Tests emotional transmission and earnedness | Affect is flat, forced, or melodramatic |
| Character Architecture | Deep psychology engine and arc mechanics | Character design/arc needs full structural drilldown |
| Interiority Preservation | Tracks POV interiority in high-intensity beats | Action/crisis scenes lose inner life |
| Female Interiority | Tracks gendered interiority thinning patterns | Significant female characters feel externally rendered |
| Literary Craft Deep Dive | Tests load-bearing literary technique vs ornament | Literary/upmarket intent across any genre |
| Force Architecture | Tests legibility, causality, cost, and meaning in physical conflict | Fights/action read as spectacle loops |
| AI-Prose Calibration | Diagnoses AI-patterned prose failure modes | Manuscript is AI-assisted or "sounds AI" |
| Dialectical Clarity | Tests argument structure and fairness | Idea-heavy or philosophical manuscripts |
| Banister (Epistemic Humility) | Tests rhetorical humility and anti-strawman integrity | Moral certainty or authorial overreach concerns |
| Shelf Positioning | Tests market shelf fit and signal clarity | Positioning, pitching, or comp selection stage |
| Series & Composite Novel | Tests multi-part coherence and standalone function | Series entries or linked-story books |
| Short Fiction | Tests compression, effect unity, ending force | Short stories, flash, novella-length compression |
| Compression | Diagnoses expendable material across 5 channels (16 flags, 4 hard gates) | Manuscript runs long, scenes feel padded, or proportional imbalance detected |
| Reception Risk | Flags cultural, political, and ideological reception risk for human sensitivity review | Author requests sensitivity pre-read, manuscript handles identity/power/trauma, or "how will this land?" questions |
| Content Advisory | Maps where the manuscript depicts intense material, at what intensity, on- or off-page | Content warnings, retailer/metadata classification, or a sensitivity-review handoff is needed |
| Narrative-Decision (StoryScope) | Structure-level AI-tell scoring for how a story is built — 30 core narrative-decision features / 33 signals (Russell et al. 2026) across 10 NarraBench dimensions and 7 bundles, per-signal contributions vs. paper-anchored human/AI means (consumer of SETEC's `narrative_decision_audit`); ships uncalibrated, anti-verdict | AI-assisted fiction where you want structure-level (not texture-level) evidence — the most rewrite-resistant AI-prose axis — or structural corroboration after AI-Prose Calibration fired |
| Argument-Decision (ArgScope) | Structure-level AI tells for how an argument is built — B1 paragraph-role arc + B2 discourse-mode mix (consumer of SETEC's `argument_decision_audit`); anti-verdict, op-ed register | Op-ed/public-debate argument prose where you want collapse-tells that survive sentence-level rewriting, or a descriptive pre-flag before a dialectical-clarity pass |
| POV Voice Profile | Profiles per-POV voice signatures and distance discipline (opt-in) | Multi-POV works where voices blur or distance drifts |
| Reader-Persona Simulation | Runs the reader-dynamics lens through several declared reading dispositions and reports where audiences diverge | You want to see how different reader segments experience the same draft |
| Idiolect Preservation | Advisory: surfaces the author's idiosyncratic voice markers so revision preserves them | Author worries edits will sand off their distinctive voice |
| Punctuation Cadence | Advisory: maps punctuation rhythm and breath as a deliberate prosody signal | Line-level rhythm or pointing feels off, mechanical, or homogenized |

**Nonfiction Argument companions** (load with the Argument Engine, not the core fiction passes):

| Companion | What it does | Use it when |
|---|---|---|
| Dialectical Clarity | Tests argument structure, thesis-antithesis balance, and fairness | Idea-heavy or argument-shaped nonfiction |
| Argument Evidence | Audits the evidence base under the claim ladder | Claims outrun their support or sourcing is thin |
| Argument Persuasion | Tests audience fit, rhetorical force, and concession discipline | The argument is sound but not landing for its readers |
| Argument Red Team | Stress-tests the argument against its strongest opposition | Before publication/testimony where the case will be attacked |
| Adversarial Evidence Review | Pressure-tests evidence under formal adversarial protocols (ACH, cross-exam, severe testing) | After Citation Verifier & Field Recon when evidence defensibility is critical |

**Prerequisite:** Dialectical Clarity runs first and writes `Argument_State.md`. Red Team, Persuasion, Evidence, and Adversarial Evidence Review read it and refuse to run without it. Adversarial Evidence Review also needs the Argument Evidence, Citation Verifier, and Field Recon outputs for the same draft, and a fresh session for high-consequence work. `apodictic-pre-writing` can seed `Argument_State.md` before a draft exists.

**By nonfiction form** (canonical: `skills/core-editor/references/nonfiction-intake-routing.md` §Default activation by form and `argument-audits-routing.md`):

| Form | Start with | Then | Research before synthesis |
|---|---|---|---|
| Op-ed, persuasive essay, open letter | Dialectical Clarity | Red Team or Persuasion | Field Recon (recommended) |
| Policy brief, recommendation memo, white paper | Dialectical Clarity | Red Team, then Evidence | Field Recon (recommended) |
| Testimony, expert affidavit | Dialectical Clarity | Red Team, then Adversarial Evidence Review | Citation Verifier and Field Recon (required) |
| Legal brief, regulatory comment | Dialectical Clarity | Red Team, then Adversarial Evidence Review | Citation Verifier and Field Recon (required) |
| Academic argument, review essay | Dialectical Clarity | Evidence | Field Recon (recommended; required for peer-reviewed publication) |
| Lens or exploratory essay (writer-confirmed) | Dialectical Clarity, `register=generative` | Red Team on cash-out points only | Optional |
| Reported feature, memoir, personal essay | Narrative Nonfiction Craft or Memoir & CNF | Dialectical Clarity when a claim burden dominates | Fact-check as needed |

---

## 6) Structural Modules (Plot and Pathway)

Use when the manuscript needs spine-level diagnosis or lacks narrative shape.

| Module | What it does | Use it when |
|---|---|---|
| Plot Architecture | Diagnoses structural spine across 50 spines in 12 families | Structural diagnosis, spine identification, pacing problems |
| Plot Selection & Coaching | Upstream structure selection, hybrid design, stuck-draft triage | Pre-drafting structure choice, stuck draft, combining structures |
| Fantasy & Series Architecture | Fantasy-specific spines (5) and series-level architectural patterns (6 + 3 rhythm patterns) | Fantasy series, multi-part fantasy, series-level coherence |
| Franklin Pathway | Pre-spine viability gate; classifies material as story-shaped, storyable, argument-with-narrative, or not storyable | Passes detect no functional story spine; nonfiction reporting pile; notes without narrative shape |

---

## 7) Specialized Audits (Genre and Mode)

Use when genre-specific architecture is central.

| Audit | What it does | Use it when |
|---|---|---|
| Mystery/Thriller Architecture | Tests clue economy, red herrings, clocks, fairness | Mystery or thriller projects |
| Horror Craft Integration | Tests dread production, destabilization, consequence embodiment | Horror or horror-hybrid projects |
| Supernatural Horror | Tests whether the supernatural apparatus produces violated-reality dread vs. supernatural-coded set dressing | Ghosts, demons, curses, hauntings, or occult systems are load-bearing |
| Grimdark / Dark Fantasy | Tests whether the grimdark apparatus produces productive darkness (moral pressure, consequential violence) vs. grimdark-coded atmosphere | Grimdark or dark-fantasy positioning, tone, or bleakness as argument |
| SFF Worldbuilding Integration | Tests whether worldbuilding carries narrative load | Science fiction or fantasy projects |
| Memoir & Creative Nonfiction | Tests truth-craft, narrator stance, ethical framing | Memoir or personal narrative projects |
| Narrative Nonfiction Craft | Tests scene/reporting integration and narrative pressure | Reported nonfiction with narrative goals |
| Historical Fiction | Tests period authenticity and attitude calibration | Historical settings and period voice |
| Comedy & Satire | Tests humor landing and tonal control | Comedy-forward or satirical manuscripts |
| Fan Fiction Conversion | Tests independence from source-IP scaffolding | Fanfic to original adaptation workflows |

---

## 8) Tag Audits (Cross-Genre Experience Layers)

Use these when the manuscript promises a specific reading experience tag.

| Tag Audit | What it does | Use it when |
|---|---|---|
| Cozy Tag | Tests safety envelope, belonging, recovery rhythm | "Cozy" is in positioning, tone, or reader promise |
| Philosophical Tag | Tests question architecture and dramatic embodiment of ideas | "Novel of ideas" or philosophical intent is explicit |
| Erotic Content | Tests intimate-scene narrative function and escalation quality | Significant erotic/intimate content is present |
| Consent Complexity | Tests power, boundary, and consent architecture | Consent dynamics are central to the narrative |
| Queer Romance/Erotica | Tests queer-specific relational/intimacy signaling and integrity | Queer romance or erotica is core to the project |

---

## 9) Research Modes (Internet-Assisted)

Use when external validation is required.

| Mode | What it does | Use it when |
|---|---|---|
| Citation Verifier | Verifies source existence, accuracy, and fit against attached claims (produces `Citation_Ledger.md`) | Nonfiction or sourced fiction headed for publication |
| Comp Validation | Checks comp recency, category fit, and viability | Querying or submission prep |
| Factual Verification | Spot-checks external claims for accuracy | Real-world facts affect trust or risk |
| Field Reconnaissance | Scouts for counterevidence, literature gaps, and source-ecosystem health (produces `Field_Reconnaissance_Report.md`) | Argument-shaped work that must survive adversarial scrutiny |
| Genre Contract Currency | Checks whether genre expectations are current | Fast-moving subgenre uncertainty |
| Representation Context | Surfaces current discourse for outside-experience writing | Sensitivity/context checks are needed |

---

## 10) Recommended Bundles

Use these as default combinations.

| Scenario | Recommended bundle |
|---|---|
| First serious diagnosis on a full draft | Core DE + Stakes System + Decision Pressure + Scene Turn |
| "Good draft, not sure why it isn't landing" | Full DE + Emotional Craft + Scene Turn |
| Submission readiness check | Core or Full DE + Pass 11 (A/B/C/D) + Shelf Positioning + Comp Validation |
| Mystery/thriller not clicking | Core DE + Mystery/Thriller Architecture + Decision Pressure + Stakes System |
| AI-assisted draft cleanup | Core DE + AI-Prose Calibration + Emotional Craft + Decision Pressure |
| Action-heavy manuscript | Core DE + Force Architecture + Stakes System + Decision Pressure |
| Literary/upmarket with thematic ambition | Core or Full DE + Literary Craft + Dialectical Clarity + Banister |
| Cozy or philosophical positioning claim | Core DE + relevant Tag Audit + Shelf Positioning |
| Horror (psychological) | Core DE + Horror Craft + Stakes System + Emotional Craft |
| Memoir or creative nonfiction | Core DE + Memoir/CNF (Gornick Layer) + Franklin Pathway (if spine uncertain) |
| Historical fiction | Core DE + Historical Fiction + Decision Pressure + Stakes System |
| Fantasy series | Core or Full DE + SFF Worldbuilding + Fantasy & Series Architecture + Series/Composite Novel |
| Queer romance | Core DE + Queer Romance/Erotica + Emotional Craft + Interiority Preservation |
| Stuck draft / structural confusion | Plot Selection & Coaching (triage) + Plot Architecture (diagnosis) |
| Nonfiction without narrative shape | Franklin Pathway → then Core DE if material is classified as storyable |
| Sensitive / legally exposed content (`constraint:risk`) | Core DE + Legal Risk Register — flags defamation, privacy, and rights-clearance exposure and routes serious items to counsel; not legal advice |
| Returning with beta-reader / critique feedback | Feedback Triage → Core DE — sort, validate, and prioritize notes before spending revision time |
| Designing a beta read worth running | Beta-Reader Instrument — turn the diagnosis's open uncertainties into a focused, non-leading reader questionnaire |
| Nonfiction argument that must hold up | Nonfiction Argument Engine (Dialectical Clarity + Argument companions) + Citation Verifier + Field Reconnaissance |

---

## 11) Minimal Rule of Thumb

If uncertain, start with:

1. **Core DE**
2. **Stakes System** (universal pressure architecture diagnostic)
3. **Decision Pressure** (universal choice plausibility diagnostic)
4. **Scene Turn** (universal scene-level mechanics diagnostic)

These three audits are recommended for every manuscript regardless of genre. Then add one specialized audit based on the loudest failure signal.