[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6DBZPNGG6SGGNF7Q4KNZGQG":3,"skill-same:01M6DBZPNGG6SGGNF7Q4KNZGQG":123},{"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":33,"linkOnly":34,"markdownContent":35,"markdownBody":36,"frontMatter":37,"refName":38,"commitSha":39,"skillDirectory":40,"forksCount":14,"repositoryHtmlUrl":41,"repositoryHomepageUrl":6,"ownerHtmlUrl":42,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":43,"install":44,"resources":57},"01M6DBZPNGG6SGGNF7Q4KNZGQG","source-command-self-reflect",null,"Extract learnings from recent PR reviews, conversations, and session patterns to update the knowledge base","Prekzursil\u002Fagent-skills-toolchain","Prekzursil","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F54636077?v=4","01M6DACZ1PCQHD99SCR1YXG28B","Andrei Mihai Visalon","01M6DACZ670WZF535V3WYYSS4Z",0,"MIT","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fsource-command-self-reflect\u002FSKILL.md",[],[19,20,21,22,23,24,25,26,27,28,29,30,31,32],"reflect","command","self reflect","self","source command self reflect","pr","update","knowledge","extract","session","patterns","learnings","base","knowledge base","2026-11-10 10:12:30",false,"---\nname: \"source-command-self-reflect\"\ndescription: \"Extract learnings from recent PR reviews, conversations, and session patterns to update the knowledge base\"\n---\n\n# source-command-self-reflect\n\nUse this skill when the user asks to run the migrated source command `self-reflect`.\n\n## Command Template\n\n# BEADS Self-Reflect\n\nYou are performing a self-reflection for the BEADS agent swarm. Your job is to analyze PR review comments, conversation history, and session patterns to extract high-quality, reusable learnings.\n\n**Philosophy**: Be judicious. Quality over quantity. Each learning should make future development measurably better.\n\n## Phase A: PR Comment Analysis\n\n### Step 1: Fetch PR Comments\n\n```bash\nGITHUB_TOKEN=$(gh auth token) npx tsx scripts\u002Fbeads-fetch-pr-comments.ts --days 7\n```\n\nThis outputs PR comments to `.beads\u002Ftemp\u002Fpr-comments.json`.\n\n### Step 2: Extract CodeRabbit's Structured Learnings\n\n```bash\ncat .beads\u002Ftemp\u002Fpr-comments.json | jq -r '.comments[].body' | grep -A5 \"^Learnt from:\" | grep \"^Learning:\" | sed 's\u002F^Learning: \u002F\u002F' | sort -u\n```\n\n### Evaluate Each CodeRabbit Learning\n\n**NOT all CodeRabbit learnings are equal.** Evaluate each one:\n\n#### ACCEPT (High Value)\n\n- **Applies to patterns**: `Applies to **\u002F*.test.ts: ...` - Actionable file-scoped rules\n- **NEVER\u002FALWAYS rules**: Clear, enforceable constraints\n- **Security\u002FPerformance**: Critical quality gates\n- **Gotchas with context**: Explains WHY something is problematic\n\n#### REJECT or DEFER (Low Value)\n\n- **PR-specific context**: \"In PR #X, someone did Y\" - Too specific unless the pattern generalizes\n- **Personality observations**: \"Developer provides detailed updates\" - Not actionable\n- **Process descriptions**: Meta, not code\n- **Duplicate with slight rewording**: Check if we already have this fact\n- **Obvious\u002Ftrivial**: Things any developer should know\n\n#### TRANSFORM (Medium Value -> Make High Value)\n\n- **Before**: \"In PR #593, developer explains optional AIProvider methods support backward compatibility\"\n- **After**: \"Optional interface methods (generateWithTools?, generateObject?) follow Interface Segregation Principle - add runtime guards before use\"\n\n### Quality Filter Questions\n\nFor each potential learning, ask:\n\n1. **Would this prevent a bug?** If yes, high priority.\n2. **Would this save review cycles?** If yes, add it.\n3. **Is this codebase-specific or universal?** Tag accordingly.\n4. **Do we already have this?** Check for semantic duplicates.\n5. **Can an agent act on this?** If not actionable, skip it.\n6. **Would this confuse a naive agent?** If a future agent without context wouldn't benefit, skip it.\n\n## Phase B: Conversation & Session Mining (Optional)\n\nAnalyze current context window and optionally historical sessions for implicit insights:\n\n**Strategic patterns to look for:**\n\n| Pattern                       | Usually indicates          | Value     |\n| ----------------------------- | -------------------------- | --------- |\n| \"The problem was...\"          | Debugging insight          | High      |\n| \"It turns out...\"             | Discovery moment           | High      |\n| \"We decided to...\"            | Architectural decision     | Very High |\n| \"The reason we...\"            | Rationale worth preserving | Very High |\n| \"Unlike what you'd expect...\" | Non-obvious behavior       | High      |\n| \"Never do X because...\"       | Gotcha\u002Fpitfall             | High      |\n\n**Filter**: Only capture insights that are codebase-specific, non-obvious, actionable, and durable.\n\n## Phase C: Config Reflection (Optional)\n\nReview Codex instructions and config for improvement opportunities:\n\n- `.claude\u002Fcommands\u002F*` - Are commands still accurate?\n- `AGENTS.md` - Any outdated guidance?\n- `.claude\u002Fsettings.json` - Missing permissions or tools?\n\nPresent findings interactively: explain issue -> propose change -> get feedback -> implement.\n\n## Step 3: Analyze Comment Patterns\n\nBeyond CodeRabbit's structured learnings, look for **recurring patterns** in review comments:\n\n```bash\ncat .beads\u002Ftemp\u002Fpr-comments.json | jq -r '.comments[] | select(.reviewerType == \"coderabbit\") | .body' | grep -i \"nitpick\\|issue\\|suggestion\\|consider\\|should\\|must\\|avoid\" | head -50\n```\n\n### Pattern Detection Checklist\n\nLook for these patterns across multiple PRs:\n\n1. **Repeated corrections**: Same issue flagged in multiple PRs -> Create a learning\n2. **Architectural feedback**: Comments about service design, dependencies -> Capture the principle\n3. **Testing feedback**: Mock patterns, test structure issues -> Document the pattern\n4. **Type safety issues**: Repeated type casting problems -> Create a gotcha\n5. **Performance concerns**: N+1 queries, memory issues -> Document with context\n\n## Step 4: Present Candidates for Approval\n\nPresent learnings as a numbered list for user selection:\n\n```markdown\n## Candidate Learnings\n\n1. **\u003CBrief title>** - \u003COne sentence description>\n\n2. **\u003CBrief title>** - \u003COne sentence description>\n\n3. **\u003CBrief title>** - \u003COne sentence description>\n\n---\n\nWhich numbers do you want to capture? (all \u002F 1,2,3 \u002F none)\n```\n\n**After user selects**, classify each approved learning interactively:\n\n```text\n**Fact**: [extracted core insight]\n**Type**: [pattern|gotcha|decision|api_behavior|security|performance]\n**Applies to**: [file patterns or components]\n**Confidence**: [high|medium|low]\n\nDoes this look right? (yes \u002F edit)\n```\n\n### Conflict Resolution\n\nWhen a new learning conflicts with an existing one:\n\n| Situation                              | Action                                |\n| -------------------------------------- | ------------------------------------- |\n| New learning is more specific\u002Faccurate | New supersedes old                    |\n| Both capture different valid aspects   | Keep both, add distinguishing context |\n| New learning contradicts old           | Ask user which is correct             |\n| Old learning is a subset of new        | Merge into comprehensive learning     |\n\n**The key question**: Would having both learnings confuse a naive agent, or complement each other?\n\n## Step 5: Categorize and Store\n\nFor each validated learning, add it to the appropriate knowledge base JSONL file following the schema in `.beads\u002Fknowledge\u002FREADME.md`.\n\n### Type Guide\n\n| Type           | When to Use                  | Example                                                 |\n| -------------- | ---------------------------- | ------------------------------------------------------- |\n| `pattern`      | Reusable code patterns       | \"Use mock factories for test data\"                      |\n| `gotcha`       | Common mistakes\u002Fpitfalls     | \"Truthy check fails for explicit zero values\"           |\n| `security`     | Security-sensitive patterns  | \"Always validate JWT server-side\"                       |\n| `performance`  | Performance implications     | \"Use database indexes for frequent queries\"             |\n| `decision`     | Team\u002Farchitectural decisions | \"Prefer Strategy over Template Method for AI providers\" |\n| `api_behavior` | External API quirks          | \"Prisma findMany returns [] not null\"                   |\n| `code_quirk`   | Codebase-specific oddities   | \"Thread model is for drafts only, not email threads\"    |\n\n### Canonicalization Rules\n\n1. **Remove PR references**: \"In PR #593...\" -> Remove, keep the principle\n2. **Remove names when not relevant**: Keep only if it's a team decision\n3. **Generalize file paths**: `src\u002Flib\u002Fservices\u002Ffoo.ts` -> `src\u002Flib\u002Fservices\u002F**\u002F*.ts`\n4. **Use imperative mood**: \"Consider using...\" -> \"Use...\"\n5. **Include the WHY**: \"Use X\" -> \"Use X because Y\"\n6. **Keep under 200 chars** when possible\n\n### Confidence Assessment\n\n| Level    | Criteria                                                                       |\n| -------- | ------------------------------------------------------------------------------ |\n| `high`   | Explicit rules (NEVER\u002FALWAYS), security issues, repeated pattern across 3+ PRs |\n| `medium` | Single PR observation, team preference, architectural guidance                 |\n| `low`    | Speculative, context-dependent, might not always apply                         |\n\n## Step 6: Deduplication Check\n\nBefore adding, check for semantic duplicates in existing knowledge files.\n\n## Step 7: Generate Report\n\n```markdown\n## Self-Reflection Results\n\n### Summary\n\n- PRs Analyzed: X\n- Comments Reviewed: Y\n- CodeRabbit Learnings Evaluated: Z\n  - Accepted: A\n  - Rejected (low value): B\n  - Transformed: C\n- Conversation\u002FSession Learnings: W\n- Total New Facts Added: N\n\n### Learnings Added\n\n#### High-Value Learnings\n\n1. **[type]** \u003Cdescription>\n   - Why accepted: \u003Creason>\n\n#### Rejected\u002FDeferred\n\n1. \"\u003Clearning>\" - \u003Creason for rejection>\n\n### Knowledge Base Statistics\n\n- Total facts: N\n- By type: pattern (X), decision (Y), gotcha (Z)...\n- Quality: high (A), medium (B), low (C)\n```\n\n## Anti-Patterns to Avoid\n\n1. **Quantity over quality**: Don't add 100 mediocre facts; add 10 great ones\n2. **Blindly trusting CodeRabbit**: Evaluate each learning critically\n3. **Ignoring duplicates**: Check before adding\n4. **Missing the WHY**: Facts without reasoning are less useful\n5. **Over-specificity**: \"Use X in file Y\" -> \"Use X in files matching pattern\"\n6. **Under-specificity**: \"Write good tests\" -> Not actionable, skip it\n","# source-command-self-reflect\n\nUse this skill when the user asks to run the migrated source command `self-reflect`.\n\n## Command Template\n\n# BEADS Self-Reflect\n\nYou are performing a self-reflection for the BEADS agent swarm. Your job is to analyze PR review comments, conversation history, and session patterns to extract high-quality, reusable learnings.\n\n**Philosophy**: Be judicious. Quality over quantity. Each learning should make future development measurably better.\n\n## Phase A: PR Comment Analysis\n\n### Step 1: Fetch PR Comments\n\n```bash\nGITHUB_TOKEN=$(gh auth token) npx tsx scripts\u002Fbeads-fetch-pr-comments.ts --days 7\n```\n\nThis outputs PR comments to `.beads\u002Ftemp\u002Fpr-comments.json`.\n\n### Step 2: Extract CodeRabbit's Structured Learnings\n\n```bash\ncat .beads\u002Ftemp\u002Fpr-comments.json | jq -r '.comments[].body' | grep -A5 \"^Learnt from:\" | grep \"^Learning:\" | sed 's\u002F^Learning: \u002F\u002F' | sort -u\n```\n\n### Evaluate Each CodeRabbit Learning\n\n**NOT all CodeRabbit learnings are equal.** Evaluate each one:\n\n#### ACCEPT (High Value)\n\n- **Applies to patterns**: `Applies to **\u002F*.test.ts: ...` - Actionable file-scoped rules\n- **NEVER\u002FALWAYS rules**: Clear, enforceable constraints\n- **Security\u002FPerformance**: Critical quality gates\n- **Gotchas with context**: Explains WHY something is problematic\n\n#### REJECT or DEFER (Low Value)\n\n- **PR-specific context**: \"In PR #X, someone did Y\" - Too specific unless the pattern generalizes\n- **Personality observations**: \"Developer provides detailed updates\" - Not actionable\n- **Process descriptions**: Meta, not code\n- **Duplicate with slight rewording**: Check if we already have this fact\n- **Obvious\u002Ftrivial**: Things any developer should know\n\n#### TRANSFORM (Medium Value -> Make High Value)\n\n- **Before**: \"In PR #593, developer explains optional AIProvider methods support backward compatibility\"\n- **After**: \"Optional interface methods (generateWithTools?, generateObject?) follow Interface Segregation Principle - add runtime guards before use\"\n\n### Quality Filter Questions\n\nFor each potential learning, ask:\n\n1. **Would this prevent a bug?** If yes, high priority.\n2. **Would this save review cycles?** If yes, add it.\n3. **Is this codebase-specific or universal?** Tag accordingly.\n4. **Do we already have this?** Check for semantic duplicates.\n5. **Can an agent act on this?** If not actionable, skip it.\n6. **Would this confuse a naive agent?** If a future agent without context wouldn't benefit, skip it.\n\n## Phase B: Conversation & Session Mining (Optional)\n\nAnalyze current context window and optionally historical sessions for implicit insights:\n\n**Strategic patterns to look for:**\n\n| Pattern                       | Usually indicates          | Value     |\n| ----------------------------- | -------------------------- | --------- |\n| \"The problem was...\"          | Debugging insight          | High      |\n| \"It turns out...\"             | Discovery moment           | High      |\n| \"We decided to...\"            | Architectural decision     | Very High |\n| \"The reason we...\"            | Rationale worth preserving | Very High |\n| \"Unlike what you'd expect...\" | Non-obvious behavior       | High      |\n| \"Never do X because...\"       | Gotcha\u002Fpitfall             | High      |\n\n**Filter**: Only capture insights that are codebase-specific, non-obvious, actionable, and durable.\n\n## Phase C: Config Reflection (Optional)\n\nReview Codex instructions and config for improvement opportunities:\n\n- `.claude\u002Fcommands\u002F*` - Are commands still accurate?\n- `AGENTS.md` - Any outdated guidance?\n- `.claude\u002Fsettings.json` - Missing permissions or tools?\n\nPresent findings interactively: explain issue -> propose change -> get feedback -> implement.\n\n## Step 3: Analyze Comment Patterns\n\nBeyond CodeRabbit's structured learnings, look for **recurring patterns** in review comments:\n\n```bash\ncat .beads\u002Ftemp\u002Fpr-comments.json | jq -r '.comments[] | select(.reviewerType == \"coderabbit\") | .body' | grep -i \"nitpick\\|issue\\|suggestion\\|consider\\|should\\|must\\|avoid\" | head -50\n```\n\n### Pattern Detection Checklist\n\nLook for these patterns across multiple PRs:\n\n1. **Repeated corrections**: Same issue flagged in multiple PRs -> Create a learning\n2. **Architectural feedback**: Comments about service design, dependencies -> Capture the principle\n3. **Testing feedback**: Mock patterns, test structure issues -> Document the pattern\n4. **Type safety issues**: Repeated type casting problems -> Create a gotcha\n5. **Performance concerns**: N+1 queries, memory issues -> Document with context\n\n## Step 4: Present Candidates for Approval\n\nPresent learnings as a numbered list for user selection:\n\n```markdown\n## Candidate Learnings\n\n1. **\u003CBrief title>** - \u003COne sentence description>\n\n2. **\u003CBrief title>** - \u003COne sentence description>\n\n3. **\u003CBrief title>** - \u003COne sentence description>\n\n---\n\nWhich numbers do you want to capture? (all \u002F 1,2,3 \u002F none)\n```\n\n**After user selects**, classify each approved learning interactively:\n\n```text\n**Fact**: [extracted core insight]\n**Type**: [pattern|gotcha|decision|api_behavior|security|performance]\n**Applies to**: [file patterns or components]\n**Confidence**: [high|medium|low]\n\nDoes this look right? (yes \u002F edit)\n```\n\n### Conflict Resolution\n\nWhen a new learning conflicts with an existing one:\n\n| Situation                              | Action                                |\n| -------------------------------------- | ------------------------------------- |\n| New learning is more specific\u002Faccurate | New supersedes old                    |\n| Both capture different valid aspects   | Keep both, add distinguishing context |\n| New learning contradicts old           | Ask user which is correct             |\n| Old learning is a subset of new        | Merge into comprehensive learning     |\n\n**The key question**: Would having both learnings confuse a naive agent, or complement each other?\n\n## Step 5: Categorize and Store\n\nFor each validated learning, add it to the appropriate knowledge base JSONL file following the schema in `.beads\u002Fknowledge\u002FREADME.md`.\n\n### Type Guide\n\n| Type           | When to Use                  | Example                                                 |\n| -------------- | ---------------------------- | ------------------------------------------------------- |\n| `pattern`      | Reusable code patterns       | \"Use mock factories for test data\"                      |\n| `gotcha`       | Common mistakes\u002Fpitfalls     | \"Truthy check fails for explicit zero values\"           |\n| `security`     | Security-sensitive patterns  | \"Always validate JWT server-side\"                       |\n| `performance`  | Performance implications     | \"Use database indexes for frequent queries\"             |\n| `decision`     | Team\u002Farchitectural decisions | \"Prefer Strategy over Template Method for AI providers\" |\n| `api_behavior` | External API quirks          | \"Prisma findMany returns [] not null\"                   |\n| `code_quirk`   | Codebase-specific oddities   | \"Thread model is for drafts only, not email threads\"    |\n\n### Canonicalization Rules\n\n1. **Remove PR references**: \"In PR #593...\" -> Remove, keep the principle\n2. **Remove names when not relevant**: Keep only if it's a team decision\n3. **Generalize file paths**: `src\u002Flib\u002Fservices\u002Ffoo.ts` -> `src\u002Flib\u002Fservices\u002F**\u002F*.ts`\n4. **Use imperative mood**: \"Consider using...\" -> \"Use...\"\n5. **Include the WHY**: \"Use X\" -> \"Use X because Y\"\n6. **Keep under 200 chars** when possible\n\n### Confidence Assessment\n\n| Level    | Criteria                                                                       |\n| -------- | ------------------------------------------------------------------------------ |\n| `high`   | Explicit rules (NEVER\u002FALWAYS), security issues, repeated pattern across 3+ PRs |\n| `medium` | Single PR observation, team preference, architectural guidance                 |\n| `low`    | Speculative, context-dependent, might not always apply                         |\n\n## Step 6: Deduplication Check\n\nBefore adding, check for semantic duplicates in existing knowledge files.\n\n## Step 7: Generate Report\n\n```markdown\n## Self-Reflection Results\n\n### Summary\n\n- PRs Analyzed: X\n- Comments Reviewed: Y\n- CodeRabbit Learnings Evaluated: Z\n  - Accepted: A\n  - Rejected (low value): B\n  - Transformed: C\n- Conversation\u002FSession Learnings: W\n- Total New Facts Added: N\n\n### Learnings Added\n\n#### High-Value Learnings\n\n1. **[type]** \u003Cdescription>\n   - Why accepted: \u003Creason>\n\n#### Rejected\u002FDeferred\n\n1. \"\u003Clearning>\" - \u003Creason for rejection>\n\n### Knowledge Base Statistics\n\n- Total facts: N\n- By type: pattern (X), decision (Y), gotcha (Z)...\n- Quality: high (A), medium (B), low (C)\n```\n\n## Anti-Patterns to Avoid\n\n1. **Quantity over quality**: Don't add 100 mediocre facts; add 10 great ones\n2. **Blindly trusting CodeRabbit**: Evaluate each learning critically\n3. **Ignoring duplicates**: Check before adding\n4. **Missing the WHY**: Facts without reasoning are less useful\n5. **Over-specificity**: \"Use X in file Y\" -> \"Use X in files matching pattern\"\n6. **Under-specificity**: \"Write good tests\" -> Not actionable, skip it\n",{"name":5,"description":7},"main","a298c929b6cbd662e1fce54302ce520d5ceea1a5",".agents\u002Fskills\u002Fsource-command-self-reflect","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain","https:\u002F\u002Fgithub.com\u002FPrekzursil","2026-11-02 14:30:57",{"source":8,"skillName":5,"commands":45,"prompt":55,"zipAvailable":56},[46,49,52],{"tool":47,"command":48},"npx","npx skills add Prekzursil\u002Fagent-skills-toolchain --skill source-command-self-reflect",{"tool":50,"command":51},"bunx","bunx skills add Prekzursil\u002Fagent-skills-toolchain --skill source-command-self-reflect",{"tool":53,"command":54},"pnpm","pnpm dlx skills add Prekzursil\u002Fagent-skills-toolchain --skill source-command-self-reflect","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fsource-command-self-reflect\u002FSKILL.md",true,[58,64,68,72,76,80,84,88,92,96,100,104,107,111,115,119],{"path":59,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":62,"sourceHtmlUrl":63},"references\u002Faspire-functions-secrets.md","REFERENCE","参考文档",3668,"https:\u002F\u002Fgithub.com\u002Fagency-black\u002FSkills\u002Fblob\u002Fmain\u002FCopia de azure-validate\u002Freferences\u002Faspire-functions-secrets.md",{"path":65,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":66,"sourceHtmlUrl":67},"references\u002Fglobal-rules.md",1215,"https:\u002F\u002Fgithub.com\u002Fagency-black\u002FSkills\u002Fblob\u002Fmain\u002FCopia de azure-validate\u002Freferences\u002Fglobal-rules.md",{"path":69,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":70,"sourceHtmlUrl":71},"references\u002Fpolicy-validation.md",1493,"https:\u002F\u002Fgithub.com\u002Fagency-black\u002FSkills\u002Fblob\u002Fmain\u002FCopia de azure-validate\u002Freferences\u002Fpolicy-validation.md",{"path":73,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":74,"sourceHtmlUrl":75},"references\u002Frecipes\u002Fazcli\u002Ferrors.md",505,"https:\u002F\u002Fgithub.com\u002Fagency-black\u002FSkills\u002Fblob\u002Fmain\u002FCopia de azure-validate\u002Freferences\u002Frecipes\u002Fazcli\u002Ferrors.md",{"path":77,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":78,"sourceHtmlUrl":79},"references\u002Frecipes\u002Fazcli\u002FREADME.md",2268,"https:\u002F\u002Fgithub.com\u002Fagency-black\u002FSkills\u002Fblob\u002Fmain\u002FCopia de azure-validate\u002Freferences\u002Frecipes\u002Fazcli\u002FREADME.md",{"path":81,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":82,"sourceHtmlUrl":83},"references\u002Frecipes\u002Fazd\u002Faspire.md",3000,"https:\u002F\u002Fgithub.com\u002Fagency-black\u002FSkills\u002Fblob\u002Fmain\u002FCopia de azure-validate\u002Freferences\u002Frecipes\u002Fazd\u002Faspire.md",{"path":85,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":86,"sourceHtmlUrl":87},"references\u002Frecipes\u002Fazd\u002Fenvironment.md",1967,"https:\u002F\u002Fgithub.com\u002Fagency-black\u002FSkills\u002Fblob\u002Fmain\u002FCopia de azure-validate\u002Freferences\u002Frecipes\u002Fazd\u002Fenvironment.md",{"path":89,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":90,"sourceHtmlUrl":91},"references\u002Frecipes\u002Fazd\u002Ferrors.md",2644,"https:\u002F\u002Fgithub.com\u002Fagency-black\u002FSkills\u002Fblob\u002Fmain\u002FCopia de azure-validate\u002Freferences\u002Frecipes\u002Fazd\u002Ferrors.md",{"path":93,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":94,"sourceHtmlUrl":95},"references\u002Frecipes\u002Fazd\u002FREADME.md",5410,"https:\u002F\u002Fgithub.com\u002Fagency-black\u002FSkills\u002Fblob\u002Fmain\u002FCopia de azure-validate\u002Freferences\u002Frecipes\u002Fazd\u002FREADME.md",{"path":97,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":98,"sourceHtmlUrl":99},"references\u002Frecipes\u002Fbicep\u002Ferrors.md",374,"https:\u002F\u002Fgithub.com\u002Fagency-black\u002FSkills\u002Fblob\u002Fmain\u002FCopia de azure-validate\u002Freferences\u002Frecipes\u002Fbicep\u002Ferrors.md",{"path":101,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":102,"sourceHtmlUrl":103},"references\u002Frecipes\u002Fbicep\u002FREADME.md",1958,"https:\u002F\u002Fgithub.com\u002Fagency-black\u002FSkills\u002Fblob\u002Fmain\u002FCopia de azure-validate\u002Freferences\u002Frecipes\u002Fbicep\u002FREADME.md",{"path":105,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":98,"sourceHtmlUrl":106},"references\u002Frecipes\u002FREADME.md","https:\u002F\u002Fgithub.com\u002Fagency-black\u002FSkills\u002Fblob\u002Fmain\u002FCopia de azure-validate\u002Freferences\u002Frecipes\u002FREADME.md",{"path":108,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":109,"sourceHtmlUrl":110},"references\u002Frecipes\u002Fterraform\u002Ferrors.md",336,"https:\u002F\u002Fgithub.com\u002Fagency-black\u002FSkills\u002Fblob\u002Fmain\u002FCopia de azure-validate\u002Freferences\u002Frecipes\u002Fterraform\u002Ferrors.md",{"path":112,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":113,"sourceHtmlUrl":114},"references\u002Frecipes\u002Fterraform\u002FREADME.md",3420,"https:\u002F\u002Fgithub.com\u002Fagency-black\u002FSkills\u002Fblob\u002Fmain\u002FCopia de azure-validate\u002Freferences\u002Frecipes\u002Fterraform\u002FREADME.md",{"path":116,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":117,"sourceHtmlUrl":118},"references\u002Fregion-availability.md",3031,"https:\u002F\u002Fgithub.com\u002Fagency-black\u002FSkills\u002Fblob\u002Fmain\u002FCopia de azure-validate\u002Freferences\u002Fregion-availability.md",{"path":120,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":121,"sourceHtmlUrl":122},"references\u002Frole-verification.md",3947,"https:\u002F\u002Fgithub.com\u002Fagency-black\u002FSkills\u002Fblob\u002Fmain\u002FCopia de azure-validate\u002Freferences\u002Frole-verification.md",[124,146,177,204,218,234],{"skillNo":125,"name":126,"nameZh":6,"description":127,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":128,"categories":129,"keywords":130,"publishedAt":33},"01M6DF7D7478R55P46REYH4QAC","clarify-requirements","Resolve genuine pre-execution ambiguity by scanning repo context, generating 1-6 grounded options, and asking 1-3 focused questions via AskUserQuestion. NOT for mid-task use (injection risk); NOT when the goal is already unambiguous; use to disambiguate vague prompts before any file edits.","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Fworkspace-hooks\u002Fskills\u002Fclarify-requirements\u002FSKILL.md",[],[131,132,133,134,135,136,137,138,139,140,141,142,143,144,145],"clarify requirements","clarify","requirements","risk","options","task","prompts","goal","scanning","use","context","repo","injection","execution","file",{"skillNo":147,"name":148,"nameZh":6,"description":149,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":150,"categories":151,"keywords":157,"publishedAt":33},"01M6DF79CA4BHJS66J8EZDBPWJ","openmm-md","Run FREE, open-source, Windows-native GPU molecular dynamics via the openmm-md MCP (OpenMM + CUDA on the local NVIDIA GPU). Use for protein \u002F protein-ligand MD when you want a free stack or are on Windows without a Schrodinger license. Covers the full workflow (prepare, solvate, minimize, equilibrate, production, analyze, mmgbsa) AND an honest decision guide vs Schrodinger Desmond, including how to run Desmond legitimately on Linux or a cloud GPU with a real license. This is a deliberately LESSER free alternative and does NOT replace the schrodinger MCP tools. Do NOT use it to pick a free tool when a validated FEP+, WaterMap, IFD-MD, or Prime MM-GBSA result is actually required.","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Fmcp\u002Fopenmm-md-mcp\u002Fskill\u002Fopenmm-md\u002FSKILL.md",[152],{"categoryNo":153,"categoryName":154,"taxonomyType":155,"taxonomyTypeDesc":156},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[158,159,160,161,162,163,164,165,166,167,168,169,170,171,140,172,173,174,175,176],"openmm md","how","decision","pick","linux","analyze","tool","stack","open","NVIDIA","gpu","license","workflow","windows","free","molecular dynamics","local","do","MCP",{"skillNo":178,"name":179,"nameZh":6,"description":180,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":181,"categories":182,"keywords":189,"publishedAt":33},"01M6DF76C1MG85VEWP8VZCBNMN","visual-review","Take screenshots of web pages and UI using Playwright for visual review and iteration","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Feditor\u002Fmetaswarm\u002Fskills\u002Fvisual-review\u002FSKILL.md",[183,186],{"categoryNo":184,"categoryName":185,"taxonomyType":155,"taxonomyTypeDesc":156},"01JBSEEDCAT000000000000003","演示文稿",{"categoryNo":187,"categoryName":188,"taxonomyType":155,"taxonomyTypeDesc":156},"01JBSEEDCAT000000000000008","测试与质量",[190,191,192,193,194,195,196,197,198,199,159,200,201,202,203],"review","visual","visual review","pages","web","false","iteration","take","check","slides","ui","screenshot","playwright","screenshots",{"skillNo":205,"name":206,"nameZh":6,"description":207,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":208,"categories":209,"keywords":210,"publishedAt":33},"01M6DF73AQVCBZKEN0YXEX9FXP","status","Diagnostic status report — shows metaswarm installation state, project setup, and potential issues","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Feditor\u002Fmetaswarm\u002Fskills\u002Fstatus\u002FSKILL.md",[],[206,211,212,213,214,215,216,217],"setup","report","project","installation","project setup","issues","state",{"skillNo":219,"name":220,"nameZh":6,"description":221,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":222,"categories":223,"keywords":225,"publishedAt":33},"01M6DF6X6C2XB0P6GX76P7TTYV","start","Use when starting work on any task, when the user mentions metaswarm, or when the user wants to begin tracked development work","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Feditor\u002Fmetaswarm\u002Fskills\u002Fstart\u002FSKILL.md",[224],{"categoryNo":153,"categoryName":154,"taxonomyType":155,"taxonomyTypeDesc":156},[220,226,136,227,228,229,140,230,231,232,233],"start issue","issue","ready","work on issue","user","work","work on","development",{"skillNo":235,"name":211,"nameZh":6,"description":236,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":237,"categories":238,"keywords":239,"publishedAt":33},"01M6DF6QMJ9AJKFKHHKRT005PQ","Interactive project setup — detects your project, configures metaswarm, writes project-local files","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Feditor\u002Fmetaswarm\u002Fskills\u002Fsetup\u002FSKILL.md",[],[211,213,174,240,215,241],"files","interactive"]