# Dependencies # Match BOTH the directory and a stray `node_modules` symlink — a # self-referential symlink was committed once (node_modules -> itself, ELOOP), # which broke binary resolution in CI + every fresh worktree. The no-slash form # also ignores a symlink so `git add -A` can never re-introduce it. node_modules node_modules/ .pnp .pnp.js .pnpm-debug.log* # Runtime state from scheduled-tasks framework (per-session lock) .claude/scheduled_tasks.lock # Build outputs dist/ build/ out/ .next/ .cache/ _next/ # Environment variables .env .env.local .env.development.local .env.test.local .env.production.local .npmrc # Logs logs/ *.log npm-debug.log* yarn-debug.log* yarn-error.log* pnpm-debug.log* lerna-debug.log* # IDE - VSCode .vscode/* !.vscode/settings.json !.vscode/tasks.json !.vscode/launch.json !.vscode/extensions.json *.code-workspace # IDE - JetBrains .idea/ *.iml *.iws *.ipr .idea_modules/ # Testing coverage/ .nyc_output/ .jest/ # OS .DS_Store .DS_Store? ._* .Spotlight-V100 .Trashes ehthumbs.db Thumbs.db # Temporary files *.tmp *.temp *.swp *.swo *~ # Turbo .turbo .turbo/ # Miscellaneous .vercel .netlify *.tsbuildinfo .eslintcache # Project Specific .cursor/rules/nx-rules.mdc .github/instructions/nx.instructions.md vite.config.*.timestamp* vitest.config.*.timestamp* **/test-report.junit.xml test-results/ *.junit.xml test-report.junit.xml # Antigravity Stability (High Churn & Large Artifacts) # Rationale: Prevents FSEvents buffer overflows in the monorepo. **/.nx/** **/.next/** **/.parcel-cache/** **/.vercel/** **/dist/** **/build/** **/out/** **/coverage/** **/.nyc_output/** **/test-results/** **/tmp/** **/.cache/** **/storybook-static/** **/playwright-report/** **/.lighthouseci/** **/.pnpm-store/** **/.contentlayer/** **/.eslintcache **/*.tsbuildinfo **/*.log **/graph.json **/.DS_Store **/ehthumbs.db **/Thumbs.db **/_next/** **/public/video_frames/** **/artifacts/recordings/** **/recordings/** **/tmp/** .env*.local # Local Content Staging # distribution/ is tracked intentionally — it holds public-facing project # strategy (AI discovery, competitor landscape, package registry, publishing # queue, content tracker, etc.) that benefits viewers and contributors. # Personal/career-sensitive files inside it stay local-only: distribution/EMPLOYER_SIGNAL.md distribution/EXPOSURE_AUDIT_LOG.md distribution/EXPOSURE_IMPACT_REVIEW.md **/articles/ !apps/docs/src/app/articles/ .eslintcache # Generated by benchmarks/suites/ilb-oxlint-parity/harvest-fixtures.mjs benchmarks/suites/ilb-oxlint-parity/harvested-fixtures/ benchmarks/results/ilb-oxlint-parity/.parity.oxlintrc.json # Agents Control Plane Sync (Private Data Boundary) **/.agents-data/ **/.agent-tools/ # Generated compiled files in source directories packages/*/src/**/*.js packages/*/src/**/*.js.map packages/*/src/**/*.d.ts # …except devkit's vendored AST/rule types: hand-run snapshot, checked in (scripts/vendor-estree-types.ts). !packages/eslint-devkit/src/types/estree/**/*.d.ts .review-screens/ .build-nocomments # scratch dir from benchmarks/scripts/repro-deep-chain.mjs (self-cleans, but not if interrupted) .deep-chain-repro/ # Written by scripts/ci-build.mts so post-build checks know what this shard built. .ci-built-packages.json benchmarks/.real-source-cache/ benchmarks/.corpus-gh-cache.json # Derived case artifacts — regenerated, never hand-edited. # # `RULE_CASES.json` alone was 135,000 lines, and every PR that touched a rule # carried it. Nobody can find a nine-line behaviour change in that, so the # review it was supposed to enable never happened. # # What stays in git is what cannot be derived: # benchmarks/cases/registry.json hand-authored — the accountability register # benchmarks/budgets/*.json ratchet baselines — the shrink-only contract # benchmarks/*_PHILOSOPHY.md, VERDICT_LEDGER.md, CASE_PHILOSOPHY.md written # # Regenerating is also what makes the locks honest. `sealed-vs-open-lock` read # the COMMITTED ledger, so it verified whatever was last generated rather than # what the tests currently say; it now regenerates first, and a poisoned # artifact no longer changes its verdict. # # npm run rule-cases # RULE_CASES.{json,md} # npm run case-registry # CASE_REGISTRY.md # npm run audit:spellings # SPELLING_MISSES.{json,md} # npm run probe:name-dependence benchmarks/RULE_CASES.json benchmarks/RULE_CASES.md benchmarks/SPELLING_MISSES.json benchmarks/SPELLING_MISSES.md benchmarks/NAME_DEPENDENCE.json benchmarks/CASE_REGISTRY.md benchmarks/CANDIDATES.md # The sabotage fixture written by benchmarks/__tests__/sealed-vs-open-lock.test.ts. # It is deleted in a `finally`, but a killed process leaves it behind — and it # deliberately contains a contradictory FN marker, so committing it would wedge # the ledger for everyone. Ignored so `git add -A` cannot pick it up. packages/*/src/tests/__coherence-probe.test.ts # Mutation HTML/JSON reports (scripts/mutation.mts) — CI uploads them as artifacts. benchmark-results/mutation/