[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6NWC6M82V1XDPW60BQPPV6W":3,"skill-same:01M6NWC6M82V1XDPW60BQPPV6W":61},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":34,"linkOnly":35,"markdownContent":36,"markdownBody":37,"frontMatter":38,"refName":39,"commitSha":40,"skillDirectory":41,"forksCount":14,"repositoryHtmlUrl":42,"repositoryHomepageUrl":6,"ownerHtmlUrl":43,"ownerBlogUrl":44,"ownerDescription":45,"updatedAt":46,"install":47,"resources":60},"01M6NWC6M82V1XDPW60BQPPV6W","self-healing",null,"Self-healing codebase system. Monitors runtime errors, test failures, and build breaks. Automatically diagnoses root cause, generates fix, validates fix, and applies if safe. Zero human intervention for routine failures.","rubicanjr\u002FFinCognis","rubicanjr","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F234912519?v=4","01M6NT156NND8J474G34SB3923","Rubi","01M6NT15B70RWXE1N3YV0BZY8A",0,"MIT","https:\u002F\u002Fgithub.com\u002Frubicanjr\u002FFinCognis\u002Fblob\u002Fmain\u002Fskills\u002Fself-healing\u002FSKILL.md",[],[19,20,21,22,23,24,25,26,5,27,28,29,30,31,32,33],"self","healing","self healing","safe","test","ZeRO","system","fix","build","human","runtime","errors","root","for","codebase","2026-11-10 10:12:30",false,"---\nname: self-healing\ndescription: \"Self-healing codebase system. Monitors runtime errors, test failures, and build breaks. Automatically diagnoses root cause, generates fix, validates fix, and applies if safe. Zero human intervention for routine failures.\"\n---\n\n# Self-Healing Codebase\n\nAutomated fault detection and repair. When a test fails, a build breaks, or a runtime error is logged, the healing pipeline wakes up, finds the root cause, generates a minimal fix, validates it, and applies it -- all without human intervention, as long as confidence is above the safety threshold.\n\n## Trigger Conditions\n\nThe healing pipeline activates on any of these events:\n\n| Trigger | Signal | Source |\n|---------|--------|--------|\n| Test failure | Any test exits non-zero | CI logs, `npm test` output |\n| Build break | Compiler\u002Fbundler error | CI logs, `npm run build` output |\n| Runtime error | Unhandled exception, 5xx response | Application logs, Vercel logs |\n| Type error | `tsc --noEmit` fails | Pre-commit hook, CI |\n| Lint failure | ESLint\u002FPrettier error blocking CI | CI logs |\n\n### Signals NOT Handled\n\nThe following always require human review -- do NOT auto-heal:\n\n- Security vulnerabilities (auth bypass, injection, XSS)\n- Database migrations (schema changes, data loss risk)\n- Dependency version conflicts (breaking API changes)\n- Flaky tests (non-deterministic failures -- use `replay` agent instead)\n\n## The 4-Phase Healing Pipeline\n\n```\nDETECT    ->    DIAGNOSE    ->    FIX    ->    VALIDATE\n  |                |               |               |\nTrigger        Root cause       Minimal         Re-run\nfires          identified       patch           test suite\n               confidence       applied         Pass? DONE\n               scored           to branch       Fail? ROLLBACK\n```\n\n### Phase 1: Detect\n\nCapture the full failure context:\n\n```\n- Error message (exact, not summarized)\n- Stack trace (all frames, not truncated)\n- File and line number of origin\n- Recent git changes touching that file (last 3 commits)\n- Environment (Node version, OS, CI vs local)\n```\n\nPass all context to the Diagnose phase. Incomplete context = low confidence = no auto-fix.\n\n### Phase 2: Diagnose\n\nThe `sleuth` agent performs root cause analysis:\n\n1. Read the failing file at the reported line\n2. Trace the call stack upward (max 3 frames)\n3. Check if a recent git commit introduced the regression (`git log -5 -- \u003Cfile>`)\n4. Classify the failure type:\n\n| Type | Examples | Confidence Boost |\n|------|----------|-----------------|\n| Regression | Line added in last commit causes error | +20% |\n| Null reference | `Cannot read property X of undefined` | +10% |\n| Import error | Module not found, wrong path | +15% |\n| Type mismatch | TS type error at known location | +15% |\n| Config error | Missing env var, wrong format | +10% |\n| Logic error | Wrong condition, off-by-one | 0% (harder to auto-fix) |\n\n**Confidence score** = base 50% + type bonus + evidence quality bonus (0-30%).\n\nIf final confidence \u003C 80%: log the diagnosis, notify in `thoughts\u002FHEALING.md`, stop. Do not attempt fix.\n\n### Phase 3: Fix\n\nThe `spark` agent generates the minimal fix:\n\n**Before touching any file:**\n```bash\ngit stash push -u -m \"self-healing: pre-fix stash for \u003Cerror-id>\"\n```\n\nFix rules:\n- Change the minimum number of lines necessary\n- Do not refactor surrounding code\n- Do not add features\n- Do not change function signatures\n- Add a comment: `\u002F\u002F self-healing fix: \u003Cshort reason>`\n\n**Fix size limits:**\n- Maximum 20 lines changed per fix\n- Maximum 3 files touched per fix\n- If fix requires more than this, confidence is forced to 0 -- escalate to human\n\n### Phase 4: Validate\n\nThe `verifier` agent runs the relevant test suite:\n\n```bash\n# For test failures: run the specific failing test + its file's full suite\nnpm test -- --testPathPattern=\"\u003Cfailing-file>\"\n\n# For build breaks: re-run the build\nnpm run build\n\n# For type errors: re-run type check\nnpx tsc --noEmit\n\n# For runtime errors: run integration tests if available\nnpm run test:integration\n```\n\n**Decision:**\n\n```\nValidation PASS:\n  -> git stash drop (discard the safety stash)\n  -> Commit the fix: \"fix: self-healing repair for \u003Cerror-id>\"\n  -> Log to thoughts\u002FHEALING.md\n  -> Update MTTR metric\n\nValidation FAIL:\n  -> git stash pop --index (restore pre-fix state exactly)\n  -> Log failure to thoughts\u002FHEALING.md\n  -> Escalate to human with full diagnosis report\n  -> Do NOT retry automatically (prevents loop)\n```\n\n## Safety Rules\n\n1. **Confidence gate**: Never apply a fix when confidence \u003C 80%\n2. **Scope gate**: Never touch security code, DB migrations, or auth logic\n3. **Size gate**: Never change more than 20 lines or 3 files in one heal\n4. **No retry loop**: If validation fails, stop. Human takes over.\n5. **Always stash**: Every fix attempt must be preceded by a git stash\n6. **One fix per trigger**: One error = one heal attempt. If the error recurs after a successful fix, it is a different problem -- do not re-heal automatically.\n\n## Agent Roles\n\n| Agent | Phase | Responsibility |\n|-------|-------|----------------|\n| `sleuth` | Diagnose | Root cause analysis, confidence scoring |\n| `spark` | Fix | Minimal patch generation |\n| `verifier` | Validate | Test suite execution, pass\u002Ffail decision |\n| `self-learner` | Post-heal | Record the pattern for future prevention |\n| `coroner` | Post-heal | Check if the same bug exists elsewhere in the codebase |\n\n## Healing Log Format\n\nAppend each healing event to `thoughts\u002FHEALING.md`:\n\n```markdown\n## Healing Event: 2026-04-07T10:15:00Z\n\n### Trigger\n- Type: Test failure\n- File: src\u002Fapi\u002Fusers.test.ts\n- Error: \"Cannot read properties of undefined (reading 'id')\"\n- Stack: users.test.ts:42 -> users.ts:87 -> db\u002Fquery.ts:31\n\n### Diagnosis (sleuth)\n- Root cause: db\u002Fquery.ts:31 returns `null` when no row found; callers assume non-null\n- Regression: commit abc123 (2026-04-06) removed the null guard\n- Classification: Null reference (confidence: +10%)\n- Evidence: stack trace + git blame confirm exact line + recent commit\n- Final confidence: 85%\n\n### Fix (spark)\n- File: src\u002Fapi\u002Fusers.ts line 87\n- Change: Added null check before accessing .id\n- Lines changed: 3\n- Stash: self-healing-20260407-101500\n\n### Validation (verifier)\n- Command: npm test -- --testPathPattern=\"src\u002Fapi\u002Fusers\"\n- Result: PASS (14 tests, 0 failures)\n- Decision: KEEP\n\n### Outcome\n- Status: HEALED\n- Time to repair: 4 minutes\n- MTTR contribution: recorded\n\n---\n\n## Healing Event: 2026-04-07T14:22:00Z\n\n### Trigger\n- Type: Build break\n- Error: Module '\".\u002Fconfig\"' has no exported member 'DATABASE_URL'\n\n### Diagnosis (sleuth)\n- Root cause: config.ts was refactored, export renamed to DB_URL\n- Confidence: 91%\n\n### Fix (spark)\n- File: src\u002Fapi\u002Fdb.ts line 3\n- Change: import { DB_URL as DATABASE_URL } from '.\u002Fconfig'\n- Lines changed: 1\n\n### Validation (verifier)\n- Command: npm run build\n- Result: PASS\n\n### Outcome\n- Status: HEALED\n- Time to repair: 2 minutes\n```\n\n## Metrics\n\nTrack in `thoughts\u002FHEALING.md` header section (update after each event):\n\n```markdown\n# Self-Healing Metrics (updated: 2026-04-07)\n\n| Metric | Value |\n|--------|-------|\n| Total healing events | 23 |\n| Successful auto-fixes | 19 (82.6%) |\n| Escalated to human | 4 (17.4%) |\n| Average MTTR (auto-fixed) | 3.8 minutes |\n| Average MTTR (escalated) | 47 minutes |\n| Most common trigger | Test failure (61%) |\n| Most common fix type | Null reference guard |\n```\n\n### Key Metrics Defined\n\n- **MTTR** (Mean Time To Repair): Time from trigger detection to validation PASS\n- **Auto-fix success rate**: Healed \u002F (Healed + Escalated)\n- **False fix rate**: Fixes that passed validation but broke something else (caught by post-deploy monitoring)\n\n## Scope Limits (Non-Negotiable)\n\nThe healing pipeline will NEVER auto-fix:\n\n```\n- Files matching: **\u002Fauth\u002F**, **\u002Fsecurity\u002F**, **\u002Fmiddleware\u002Fauth*\n- Files matching: **\u002Fmigrations\u002F**, **\u002Fschema.*\n- Environment variables or secrets\n- Package versions in package.json \u002F requirements.txt\n- CI\u002FCD configuration (.github\u002F**, vercel.json, Dockerfile)\n- Any file touched by a security-related commit message\n```\n\nIf the failing code is in a scoped-out path, the pipeline logs the trigger and immediately escalates with the full diagnosis -- no fix attempt.\n\n## Activation\n\nThis skill is referenced by:\n- PostToolUse hooks when test\u002Fbuild commands exit non-zero\n- The `sentinel` agent during incident response\n- The `verifier` agent when a final validation fails\n\nTo invoke manually:\n\n```\nUse self-healing to diagnose and fix the failing test in src\u002Fapi\u002Fusers.test.ts.\nError: \"Cannot read properties of undefined (reading 'id')\"\nStack trace: [paste full trace]\n```\n","# Self-Healing Codebase\n\nAutomated fault detection and repair. When a test fails, a build breaks, or a runtime error is logged, the healing pipeline wakes up, finds the root cause, generates a minimal fix, validates it, and applies it -- all without human intervention, as long as confidence is above the safety threshold.\n\n## Trigger Conditions\n\nThe healing pipeline activates on any of these events:\n\n| Trigger | Signal | Source |\n|---------|--------|--------|\n| Test failure | Any test exits non-zero | CI logs, `npm test` output |\n| Build break | Compiler\u002Fbundler error | CI logs, `npm run build` output |\n| Runtime error | Unhandled exception, 5xx response | Application logs, Vercel logs |\n| Type error | `tsc --noEmit` fails | Pre-commit hook, CI |\n| Lint failure | ESLint\u002FPrettier error blocking CI | CI logs |\n\n### Signals NOT Handled\n\nThe following always require human review -- do NOT auto-heal:\n\n- Security vulnerabilities (auth bypass, injection, XSS)\n- Database migrations (schema changes, data loss risk)\n- Dependency version conflicts (breaking API changes)\n- Flaky tests (non-deterministic failures -- use `replay` agent instead)\n\n## The 4-Phase Healing Pipeline\n\n```\nDETECT    ->    DIAGNOSE    ->    FIX    ->    VALIDATE\n  |                |               |               |\nTrigger        Root cause       Minimal         Re-run\nfires          identified       patch           test suite\n               confidence       applied         Pass? DONE\n               scored           to branch       Fail? ROLLBACK\n```\n\n### Phase 1: Detect\n\nCapture the full failure context:\n\n```\n- Error message (exact, not summarized)\n- Stack trace (all frames, not truncated)\n- File and line number of origin\n- Recent git changes touching that file (last 3 commits)\n- Environment (Node version, OS, CI vs local)\n```\n\nPass all context to the Diagnose phase. Incomplete context = low confidence = no auto-fix.\n\n### Phase 2: Diagnose\n\nThe `sleuth` agent performs root cause analysis:\n\n1. Read the failing file at the reported line\n2. Trace the call stack upward (max 3 frames)\n3. Check if a recent git commit introduced the regression (`git log -5 -- \u003Cfile>`)\n4. Classify the failure type:\n\n| Type | Examples | Confidence Boost |\n|------|----------|-----------------|\n| Regression | Line added in last commit causes error | +20% |\n| Null reference | `Cannot read property X of undefined` | +10% |\n| Import error | Module not found, wrong path | +15% |\n| Type mismatch | TS type error at known location | +15% |\n| Config error | Missing env var, wrong format | +10% |\n| Logic error | Wrong condition, off-by-one | 0% (harder to auto-fix) |\n\n**Confidence score** = base 50% + type bonus + evidence quality bonus (0-30%).\n\nIf final confidence \u003C 80%: log the diagnosis, notify in `thoughts\u002FHEALING.md`, stop. Do not attempt fix.\n\n### Phase 3: Fix\n\nThe `spark` agent generates the minimal fix:\n\n**Before touching any file:**\n```bash\ngit stash push -u -m \"self-healing: pre-fix stash for \u003Cerror-id>\"\n```\n\nFix rules:\n- Change the minimum number of lines necessary\n- Do not refactor surrounding code\n- Do not add features\n- Do not change function signatures\n- Add a comment: `\u002F\u002F self-healing fix: \u003Cshort reason>`\n\n**Fix size limits:**\n- Maximum 20 lines changed per fix\n- Maximum 3 files touched per fix\n- If fix requires more than this, confidence is forced to 0 -- escalate to human\n\n### Phase 4: Validate\n\nThe `verifier` agent runs the relevant test suite:\n\n```bash\n# For test failures: run the specific failing test + its file's full suite\nnpm test -- --testPathPattern=\"\u003Cfailing-file>\"\n\n# For build breaks: re-run the build\nnpm run build\n\n# For type errors: re-run type check\nnpx tsc --noEmit\n\n# For runtime errors: run integration tests if available\nnpm run test:integration\n```\n\n**Decision:**\n\n```\nValidation PASS:\n  -> git stash drop (discard the safety stash)\n  -> Commit the fix: \"fix: self-healing repair for \u003Cerror-id>\"\n  -> Log to thoughts\u002FHEALING.md\n  -> Update MTTR metric\n\nValidation FAIL:\n  -> git stash pop --index (restore pre-fix state exactly)\n  -> Log failure to thoughts\u002FHEALING.md\n  -> Escalate to human with full diagnosis report\n  -> Do NOT retry automatically (prevents loop)\n```\n\n## Safety Rules\n\n1. **Confidence gate**: Never apply a fix when confidence \u003C 80%\n2. **Scope gate**: Never touch security code, DB migrations, or auth logic\n3. **Size gate**: Never change more than 20 lines or 3 files in one heal\n4. **No retry loop**: If validation fails, stop. Human takes over.\n5. **Always stash**: Every fix attempt must be preceded by a git stash\n6. **One fix per trigger**: One error = one heal attempt. If the error recurs after a successful fix, it is a different problem -- do not re-heal automatically.\n\n## Agent Roles\n\n| Agent | Phase | Responsibility |\n|-------|-------|----------------|\n| `sleuth` | Diagnose | Root cause analysis, confidence scoring |\n| `spark` | Fix | Minimal patch generation |\n| `verifier` | Validate | Test suite execution, pass\u002Ffail decision |\n| `self-learner` | Post-heal | Record the pattern for future prevention |\n| `coroner` | Post-heal | Check if the same bug exists elsewhere in the codebase |\n\n## Healing Log Format\n\nAppend each healing event to `thoughts\u002FHEALING.md`:\n\n```markdown\n## Healing Event: 2026-04-07T10:15:00Z\n\n### Trigger\n- Type: Test failure\n- File: src\u002Fapi\u002Fusers.test.ts\n- Error: \"Cannot read properties of undefined (reading 'id')\"\n- Stack: users.test.ts:42 -> users.ts:87 -> db\u002Fquery.ts:31\n\n### Diagnosis (sleuth)\n- Root cause: db\u002Fquery.ts:31 returns `null` when no row found; callers assume non-null\n- Regression: commit abc123 (2026-04-06) removed the null guard\n- Classification: Null reference (confidence: +10%)\n- Evidence: stack trace + git blame confirm exact line + recent commit\n- Final confidence: 85%\n\n### Fix (spark)\n- File: src\u002Fapi\u002Fusers.ts line 87\n- Change: Added null check before accessing .id\n- Lines changed: 3\n- Stash: self-healing-20260407-101500\n\n### Validation (verifier)\n- Command: npm test -- --testPathPattern=\"src\u002Fapi\u002Fusers\"\n- Result: PASS (14 tests, 0 failures)\n- Decision: KEEP\n\n### Outcome\n- Status: HEALED\n- Time to repair: 4 minutes\n- MTTR contribution: recorded\n\n---\n\n## Healing Event: 2026-04-07T14:22:00Z\n\n### Trigger\n- Type: Build break\n- Error: Module '\".\u002Fconfig\"' has no exported member 'DATABASE_URL'\n\n### Diagnosis (sleuth)\n- Root cause: config.ts was refactored, export renamed to DB_URL\n- Confidence: 91%\n\n### Fix (spark)\n- File: src\u002Fapi\u002Fdb.ts line 3\n- Change: import { DB_URL as DATABASE_URL } from '.\u002Fconfig'\n- Lines changed: 1\n\n### Validation (verifier)\n- Command: npm run build\n- Result: PASS\n\n### Outcome\n- Status: HEALED\n- Time to repair: 2 minutes\n```\n\n## Metrics\n\nTrack in `thoughts\u002FHEALING.md` header section (update after each event):\n\n```markdown\n# Self-Healing Metrics (updated: 2026-04-07)\n\n| Metric | Value |\n|--------|-------|\n| Total healing events | 23 |\n| Successful auto-fixes | 19 (82.6%) |\n| Escalated to human | 4 (17.4%) |\n| Average MTTR (auto-fixed) | 3.8 minutes |\n| Average MTTR (escalated) | 47 minutes |\n| Most common trigger | Test failure (61%) |\n| Most common fix type | Null reference guard |\n```\n\n### Key Metrics Defined\n\n- **MTTR** (Mean Time To Repair): Time from trigger detection to validation PASS\n- **Auto-fix success rate**: Healed \u002F (Healed + Escalated)\n- **False fix rate**: Fixes that passed validation but broke something else (caught by post-deploy monitoring)\n\n## Scope Limits (Non-Negotiable)\n\nThe healing pipeline will NEVER auto-fix:\n\n```\n- Files matching: **\u002Fauth\u002F**, **\u002Fsecurity\u002F**, **\u002Fmiddleware\u002Fauth*\n- Files matching: **\u002Fmigrations\u002F**, **\u002Fschema.*\n- Environment variables or secrets\n- Package versions in package.json \u002F requirements.txt\n- CI\u002FCD configuration (.github\u002F**, vercel.json, Dockerfile)\n- Any file touched by a security-related commit message\n```\n\nIf the failing code is in a scoped-out path, the pipeline logs the trigger and immediately escalates with the full diagnosis -- no fix attempt.\n\n## Activation\n\nThis skill is referenced by:\n- PostToolUse hooks when test\u002Fbuild commands exit non-zero\n- The `sentinel` agent during incident response\n- The `verifier` agent when a final validation fails\n\nTo invoke manually:\n\n```\nUse self-healing to diagnose and fix the failing test in src\u002Fapi\u002Fusers.test.ts.\nError: \"Cannot read properties of undefined (reading 'id')\"\nStack trace: [paste full trace]\n```\n",{"name":5,"description":7},"main","70e1c350500ee750987420bbae48837afa906b7e","skills\u002Fself-healing","https:\u002F\u002Fgithub.com\u002Frubicanjr\u002FFinCognis","https:\u002F\u002Fgithub.com\u002Frubicanjr","linktr.ee\u002Ffinanskodu","Industrial Engineer & CIO @FinansKodu. Architecting AI-driven financial agents and autonomous workflows. 🏗️ Building systems, not just prompts.","2026-06-09 13:50:15",{"source":8,"skillName":5,"commands":48,"prompt":58,"zipAvailable":59},[49,52,55],{"tool":50,"command":51},"npx","npx skills add rubicanjr\u002FFinCognis --skill self-healing",{"tool":53,"command":54},"bunx","bunx skills add rubicanjr\u002FFinCognis --skill self-healing",{"tool":56,"command":57},"pnpm","pnpm dlx skills add rubicanjr\u002FFinCognis --skill self-healing","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Frubicanjr\u002FFinCognis\u002Fblob\u002Fmain\u002Fskills\u002Fself-healing\u002FSKILL.md",true,[],[62,83,92,112,141,158],{"skillNo":63,"name":64,"nameZh":6,"description":65,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":66,"categories":67,"keywords":73,"publishedAt":34},"01M6NWS9DAXJTADT4T5MB49X8W","workflow-router","Goal-based workflow orchestration - routes tasks to specialist agents based on user goals","https:\u002F\u002Fgithub.com\u002Frubicanjr\u002FFinCognis\u002Fblob\u002Fmain\u002Fskills\u002Fworkflow-router\u002FSKILL.md",[68],{"categoryNo":69,"categoryName":70,"taxonomyType":71,"taxonomyTypeDesc":72},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[74,75,76,77,78,79,80,81,82],"workflow router","workflow","router","orchestration","goals","goal","user","workflow orchestration","tasks",{"skillNo":84,"name":85,"nameZh":6,"description":86,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":87,"categories":88,"keywords":89,"publishedAt":34},"01M6NWS3RETSC0SHP16HP66E14","wiring","Wiring Verification","https:\u002F\u002Fgithub.com\u002Frubicanjr\u002FFinCognis\u002Fblob\u002Fmain\u002Fskills\u002Fwiring\u002FSKILL.md",[],[85,90,91],"verification","false",{"skillNo":93,"name":94,"nameZh":6,"description":95,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":96,"categories":97,"keywords":101,"publishedAt":34},"01M6NWS0XW9CVQZBDA1EGW0TWC","websocket-patterns","Connection management, room patterns, reconnection strategies, message buffering, and binary protocol design.","https:\u002F\u002Fgithub.com\u002Frubicanjr\u002FFinCognis\u002Fblob\u002Fmain\u002Fskills\u002Fwebsocket-patterns\u002Fskill.md",[98],{"categoryNo":99,"categoryName":100,"taxonomyType":71,"taxonomyTypeDesc":72},"01JBSEEDCAT000000000000015","视觉设计",[102,103,104,105,106,107,108,109,110,111],"websocket","patterns","websocket patterns","design","protocol","management","message","strategies","binary","connection",{"skillNo":113,"name":114,"nameZh":6,"description":115,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":116,"categories":117,"keywords":122,"publishedAt":34},"01M6NWRDQDX9ZNETAWRRMQWV52","visual-verdict","Screenshot comparison QA for frontend development. Takes a screenshot of the current implementation, scores it across multiple visual dimensions, and returns a structured PASS\u002FREVISE\u002FFAIL verdict with concrete fixes. Use when implementing UI from a design reference or verifying visual correctness.","https:\u002F\u002Fgithub.com\u002Frubicanjr\u002FFinCognis\u002Fblob\u002Fmain\u002Fskills\u002Fvisual-verdict\u002FSKILL.md",[118,121],{"categoryNo":119,"categoryName":120,"taxonomyType":71,"taxonomyTypeDesc":72},"01JBSEEDCAT000000000000009","前端开发",{"categoryNo":69,"categoryName":70,"taxonomyType":71,"taxonomyTypeDesc":72},[123,124,125,126,127,128,129,130,131,32,132,133,134,105,135,136,137,138,139,140],"visual","verdict","visual verdict","ui","across","revise","comparison","correctness","implementing","fixes","dimensions","frontend","development","qa","screenshot","returns","use","frontend development",{"skillNo":142,"name":143,"nameZh":6,"description":144,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":145,"categories":146,"keywords":147,"publishedAt":34},"01M6NWR8HSNRS7MNDK2J02QYMN","verification-loop","Comprehensive verification system covering build, types, lint, tests, security, and diff review before a PR.","https:\u002F\u002Fgithub.com\u002Frubicanjr\u002FFinCognis\u002Fblob\u002Fmain\u002Fskills\u002Fverification-loop\u002FSKILL.md",[],[148,149,90,150,151,152,153,25,154,27,155,156,157],"verification loop","loop","lint","review","security","pr","diff","tests","diff review","types",{"skillNo":159,"name":160,"nameZh":6,"description":161,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":162,"categories":163,"keywords":164,"publishedAt":34},"01M6NWR4HHSFR0VR5EXFRWA747","vector-db-patterns","Embedding strategies, ANN algorithms, hybrid search, RAG chunking strategies, and reranking for semantic search and retrieval.","https:\u002F\u002Fgithub.com\u002Frubicanjr\u002FFinCognis\u002Fblob\u002Fmain\u002Fskills\u002Fvector-db-patterns\u002Fskill.md",[],[165,166,167,168,103,169,170,171,172,109,173,174,175,176,177,32,178,179],"vector db patterns","vector db","db","vector","semantic search","search","embedding strategies","algorithms","chunking","semantic","retrieval","embedding","Hybrid Search","hybrid","rag"]