feat!: remove review-code — superseded by Quinn (2.0.0)

.github/workflows/ci.yml

@@ -31,14 +31,14 @@ jobs:
         run: |
           node --check bin/rewrite-release-notes.mjs
           node --check bin/build-updater-manifest.mjs
-          node --check bin/review-code.mjs
-          node --check lib/code-review.mjs
+          node --check bin/apply-branch-protection.mjs
+          node --check lib/branch-protection.mjs
 
       - name: Help works
         run: |
           node bin/rewrite-release-notes.mjs --help
           node bin/build-updater-manifest.mjs --help
-          node bin/review-code.mjs --help
+          node bin/apply-branch-protection.mjs --help
 
       # Smoke: dry-run renders without an LLM call. Uses the repo's own tags
       # if any exist; otherwise creates synthetic ones so the script has
@@ -87,5 +87,5 @@ jobs:
             console.log("manifest smoke ok");
           '
 
-      - name: Test reusable review tool
+      - name: Test
         run: npm test

README.md

@@ -185,57 +185,6 @@ and writes a manifest in the exact shape Tauri's updater expects. Exits
 non-zero if any binary is missing its signature, so the `publish` job fails
 loudly when signing didn't run.
 
-## Model-backed code review
-
-`review-code` is a reusable, Clawpatch-inspired review harness for protoLabs
-repos. It maps a repository into bounded feature records, asks an
-OpenAI-compatible gateway for strict JSON findings, persists those findings
-locally, and emits a Markdown report. It does not edit code.
-
-```bash
-# In the repo you want reviewed:
-npx -p @protolabsai/release-tools review-code init
-npx -p @protolabsai/release-tools review-code map
-GATEWAY_API_KEY=... OPENAI_BASE_URL=https://api.proto-labs.ai/v1 \
-  npx -p @protolabsai/release-tools review-code run --model protolabs/fast --limit 1
-npx -p @protolabsai/release-tools review-code report
-```
-
-Review state is written to `.release-tools-review/`. Add that directory to a
-repo's `.gitignore` if you use the tool regularly. The default model is
-`protolabs/smart`; override it with `--model` or `CODE_REVIEW_MODEL` when a
-gateway key is routed to another review model such as Minimax.
-
-For better feature boundaries, add a repo-local `review-code.config.json`:
-
-```json
-{
-  "features": [
-    {
-      "feature_id": "engine_core",
-      "name": "Engine core",
-      "description": "Deterministic resolver and replay-critical state.",
-      "owned_globs": ["src/engine/**/*.js"],
-      "context_globs": ["docs/engine/**/*.md"],
-      "test_globs": ["test/engine/**/*.test.js"]
-    }
-  ]
-}
-```
-
-Useful commands:
-
-```bash
-review-code status
-review-code run --feature-id engine_core
-review-code run --all
-review-code report --output ./review-report.md
-```
-
-The harness keeps prompts bounded, skips external symlinks, rejects findings
-outside the reviewed file allowlist, and preserves existing triage fields when
-a finding is regenerated.
-
 ## Branch protection defaults
 
 `apply-branch-protection` applies the protoLabs recommended branch protection
@@ -273,7 +222,6 @@ reference.
 npm install
 node bin/rewrite-release-notes.mjs --help
 node bin/build-updater-manifest.mjs --help
-node bin/review-code.mjs --help
 node bin/apply-branch-protection.mjs --help
 npm test
 ```