repo-audit-with-execution-batcheslisted

# Repo Audit with Execution Batches ## The 7-Part Audit Run in order. Each batch produces a report section. Don't batch all at once — audit reveals issues that change subsequent batch findings. ### Batch 1 — Surface Scan Same first step as `repo-discovery` — reuse that skill's Surface Scan/Entry Points steps instead of re-deriving them here. Identify: language, framework, build system, test runner, CI config. ### Batch 2 — Marker Hunt Search for: `FIX`, `PATCH`, `TODO`, `FIXME`, `HACK`, `XXX`, `BUG`, `WORKAROUND`, `temp`, `temporary`, `hack` ```bash grep -rn "FIX\|TODO\|HACK\|FIXME\|BUG\|WORKAROUND" --include="*.js" --include="*.ts" --include="*.py" <dirs> ``` Report counts per category, not every occurrence. ### Batch 3 — Doc Freshness Presence check only — does `README.md`/`AGENTS.md`/`CLAUDE.md`/`ARCHITECTURE.md`/`docs/` exist and look current at a glance? For a real claim-by-claim verification of docs against live code (routes, migration counts, design tokens, cron schedules, story IDs) plus no-duplication consolidation, that's a distinct pass — use `docs-reality-audit`, don't improvise it inline here. ### Batch 4 — Structural Validation Against `PROJECT_STRUCTURE_STANDARD.md` (from OliverCode contracts): - Does the layout match conventions? - Are there files at wrong paths? - Are there orphaned files? ### Batch 5 — Config Drift - `package.json` vs lockfile — in sync? - `.env.example`/`.env.local.example` vs actual env vars used in code - CI config vs actual