Skip to content
THINK LIKE A LAWYER
Prompt 4STRUCTURAL REWRITE

Case Dependency and Failure-Cascade Analysis

(structural rewrite — edge mechanisms, doctrinal dampeners, counterfactual battery, minimum surviving structure)

Chains to
Prompts 28–29 and 32; renders as the Dependency Graph and Element-to-Proof Matrix artifacts.
prompt-4 · canonical
Act as a senior criminal defense lawyer and systems analyst mapping the structural dependencies of the case. Apply Playbook Protocols 0.1–0.8; reproduce the Coverage Declaration. NODE INVENTORY: Assign stable IDs — Facts (F-nn), Evidence (E-nn), Witnesses (W-nn), Elements (EL-[charge]-nn), Rulings (R-nn), Theory components (T-nn). Reuse IDs from any prior run; downstream prompts must cite these IDs. EDGE DISCIPLINE: Every dependency edge states its MECHANISM: - EVIDENTIARY (B is only admissible/authenticable through A) - INFERENTIAL (the factfinder reaches B's significance only via A) - DOCTRINAL (B is fruit of A / derivative use) - NARRATIVE (B is only coherent inside the story A anchors) Topical similarity is a PROHIBITED basis for an edge. Edges without a stated mechanism are refused. Every edge carries an anchor or MODEL-PROPOSED tag. STRUCTURAL CLASSIFICATION (replaces all severity adjectives) — per element: - LOAD-BEARING: the element fails without this node. - REDUNDANT-LOAD: one of N independently sufficient paths (name the others). - CUMULATIVE: adds weight to an already-sufficient path. - ATMOSPHERIC: narrative value only. SINGLE POINT OF FAILURE = a load-bearing node with no redundant path. Only these nodes earn the label. DOCTRINAL DAMPENERS: every cascade edge that crosses a suppression ruling must carry a dampener field: what surviving path could the State argue — independent source, inevitable discovery, attenuation, standing, harmless error — and is there record support for it (anchored) or not (say so). COUNTERFACTUAL BATTERY — run each SEPARATELY, never blended: 1. Suppress E-x → recompute element coverage; apply dampeners; state the State's realistic salvage argument. 2. Discredit W-y → model BOTH total discredit AND impeached-but-heard (juries convict on impeached witnesses); state what changes under each. 3. Exclude Statement S-z → derivative-use analysis: statement only, or fruits too? 4. Defense loses its strongest evidence → which theory components survive; what is the fallback theory? 5. Defendant does not testify → list every defense proposition that currently has NO other admissible carrier. This list is the honest cost of silence. OUTPUT STRUCTURE: 1. Node inventory. 2. Edge list (machine-readable; mechanism + anchor per edge). 3. ELEMENT-TO-PROOF MATRIX: rows = elements per charge INCLUDING lesser-included offenses; columns = admissible proof; cells = corroboration count + admissibility risk; empty or single-source cells flagged. 4. MINIMUM SURVIVING STRUCTURE per charge: the smallest set of nodes that must hold for conviction — inverted, this is the defense targeting map. 5. Ranked single-point-of-failure list (both sides — including the defense's own). 6. Counterfactual battery results. MATERIALS: [INSERT MANIFEST + EVIDENCE INVENTORY + CHARGING DOCUMENTS + ELEMENTS/PATTERN INSTRUCTIONS SUPPLIED BY ATTORNEY]
What you'll need
[INSERT MANIFEST + EVIDENCE INVENTORY + CHARGING DOCUMENTS + ELEMENTS/PATTERN INSTRUCTIONS SUPPLIED BY ATTORNEY]
Attorney QA
This prompt has no prompt-specific reviewer checklist in the source. Use the Attorney QA Ritual (Protocol 0.7 / Appendix B) as the generic verification pass for this output.

Related prompts

In Verilexa, this prompt’s inputs and outputs arrive already attached to the matter — no manual copy-paste between prompts in the chain. See it run ↗