cursorqb-autopsy

# CursorQB Autopsy Inspect an existing or partially built repository and write `Planner-docs/Autopsy.md`. When enough evidence exists, also create/update `Planner-docs/Project-Ontology.md` and `Planner-docs/Project-Comprehension.md`. ## Sources - Primary: `Planner-docs/Main-Planing.md` - Optional continuity: `Planner-docs/Planing-Ledger.md` - Repo evidence: README, manifests, source/service/package dirs, tests, docs, configs, CI - Full Step 1.5 spec: `autopsy-planner.md` - Validator entrypoint: `scripts/cursorqb_launcher.py --controller planner-validator` ## Language Contract CursorQB asks intake questions in the user's language when practical. Generated Planner-docs artifacts are English by default unless the user explicitly requests another body language. Required document headings remain English for validator stability. ## Plan-Run Contract Use Cursor Plan Mode and bind this run through the bundled deterministic Plan-Run controller: - Outcome: produce `Planner-docs/Autopsy.md`; optionally produce `Planner-docs/Project-Ontology.md` and `Planner-docs/Project-Comprehension.md`. - Files allowed to modify: `Planner-docs/Autopsy.md`, `Planner-docs/Project-Ontology.md`, and `Planner-docs/Project-Comprehension.md`. - Files forbidden to modify: source, config, tests, scripts, `Planner-docs/Main-Planing.md`, Step 2 files, and Step 3 files. - Validation checkpoints: run `python3 -I -S -B <plugin-root>/scripts/cursorqb_launcher.py --controller planner-validator -- --root . --mode autopsy --strict`. - Stop gates: missing main plan, empty/new repo skip, repository read errors, unrelated dirty worktree, or secret exposure risk. - Subagent policy: use read-only subagents only for large repo exploration, ontology mapping, readiness review, or security review; parent CursorQB writes official artifacts. ## Required Behavior 1. Read `autopsy-planner.md` fully. 2. Read `Main-Planing.md`. 3. Inspect the repo with safe read-only commands. 4. Skip for empty or nearly empty repos. 5. Write `Autopsy.md` with required headings. 6. Write `Project-Ontology.md` and `Project-Comprehension.md` only when enough evidence exists. 7. Redact secret-like values and report only paths/line numbers/pattern names. ## Completion Report Report whether Step 1.5 succeeded, skipped, or blocked; whether Autopsy/Ontology changed; top findings; validator result; and how Step 2 should use the evidence.