[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6X23WTJ7233E81EEKT0DBS3":3,"skill-same:01M6X23WTJ7233E81EEKT0DBS3":73},{"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":23,"publishedAt":42,"linkOnly":43,"markdownContent":44,"markdownBody":45,"frontMatter":46,"refName":49,"commitSha":50,"skillDirectory":51,"forksCount":52,"repositoryHtmlUrl":53,"repositoryHomepageUrl":54,"ownerHtmlUrl":55,"ownerBlogUrl":56,"ownerDescription":57,"updatedAt":58,"install":59,"resources":72},"01M6X23WTJ7233E81EEKT0DBS3","reflect",null,"Learning capture system that extracts HIGH\u002FMED\u002FLOW confidence patterns from conversations to prevent repeating mistakes. Use after user corrections (\"no\", \"wrong\"), praise (\"perfect\", \"exactly\"), or when discovering edge cases. Complements .squad\u002Fagents\u002F{agent}\u002Fhistory.md and .squad\u002Fdecisions.md.","github\u002Fsquad","github","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F9919?v=4","01M6AC83DS133TPXFSP1DHNCC6","GitHub","01M6X0MZ6B0WYMMDPK3Q61KER6",3391,"MIT","https:\u002F\u002Fgithub.com\u002Fgithub\u002Fsquad\u002Fblob\u002Fdev\u002Ftemplates\u002Fskills\u002Freflect\u002FSKILL.md",[18],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[5,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41],"when","memory","mistakes","Wrong","high","conversations","patterns","use","learning capture","user","decisions","capture","history","system","learning","team","edge","squad","2026-11-10 10:12:30",false,"---\nname: reflect\ndescription: Learning capture system that extracts HIGH\u002FMED\u002FLOW confidence patterns from conversations to prevent repeating mistakes. Use after user corrections (\"no\", \"wrong\"), praise (\"perfect\", \"exactly\"), or when discovering edge cases. Complements .squad\u002Fagents\u002F{agent}\u002Fhistory.md and .squad\u002Fdecisions.md.\nlicense: MIT\nversion: 1.0.0-squad\ndomain: team-memory, learning\nconfidence: high\n---\n\n# Reflect Skill\n\n**Critical learning capture system** for Squad. Prevents repeating mistakes and preserves successful patterns across sessions.\n\nAnalyze conversations and propose improvements to squad knowledge based on what worked, what didn't, and edge cases discovered. **Every correction is a learning opportunity.**\n\n---\n\n## Integration with Squad Architecture\n\n**Reflect complements existing Squad knowledge systems:**\n\n1. **Governed runtime memory** — Durable observations and decisions persisted through configured state tools\n2. **Tracked governing source** — Coordinator prompts, skills, and other behavior-defining files changed through focused pull requests\n3. **`reflect` skill** — Identifies learnings that may warrant either runtime memory or an approved tracked behavior change\n\n**Workflow:**\n- Use `reflect` during work to capture learnings\n- At session end, review captured learnings\n- Persist evidence through governed runtime memory when available\n- Promote accepted behavior changes through the accountable specialist as tracked source changes in a focused PR\n- Report adjacent improvements separately; do not expand the current scope without explicit user approval\n\n---\n\n## Triggers\n\n### ? HIGH Priority (Invoke Immediately)\n\n| Trigger | Example | Why Critical |\n|---------|---------|--------------|\n| User correction | \"no\", \"wrong\", \"not like that\", \"never do\" | Captures mistakes to prevent repetition |\n| Architectural insight | \"you removed that without understanding why\" | Documents design decisions (Chesterton's Fence) |\n| Immediate fixes | \"debug\", \"root cause\", \"fix all\" | Learns from errors in real-time |\n\n### ? MEDIUM Priority (Invoke After Multiple)\n\n| Trigger | Example | Why Important |\n|---------|---------|---------------|\n| User praise | \"perfect\", \"exactly\", \"great\" | Reinforces successful patterns |\n| Tool preferences | \"use X instead of Y\", \"prefer\" | Builds workflow preferences |\n| Edge cases | \"what if X happens?\", \"don't forget\", \"ensure\" | Captures scenarios to handle |\n\n### ? LOW Priority (Invoke at Session End)\n\n| Trigger | Example | Why Useful |\n|---------|---------|------------|\n| Repeated patterns | Frequent use of specific commands\u002Ftools | Identifies workflow preferences |\n| Session end | After complex work | Consolidates all session learnings |\n\n---\n\n## Process\n\n### Phase 1: Identify Learning Target\n\nDetermine what knowledge system should be updated:\n\n1. **Observation or preference** → Governed runtime memory, when configured\n2. **Accepted team-wide behavior change** → The accountable tracked instruction or skill source, through a focused PR\n3. **Adjacent improvement** → Report separately and require explicit user approval before adding it to scope\n\n### Phase 2: Analyze Conversation\n\nScan for learning signals with confidence levels:\n\n#### HIGH Confidence: Corrections\n\nUser actively steered or corrected output.\n\n**Detection patterns:**\n- Explicit rejection: \"no\", \"not like that\", \"that's wrong\"\n- Strong directives: \"never do\", \"always do\", \"don't ever\"\n- User provided alternative implementation\n\n**Example:**\n```text\nUser: \"No, use the azure-devops MCP tool instead of raw API calls\"\n→ [HIGH] + Add constraint: \"Prefer azure-devops MCP tools over REST API\"\n```\n\n#### MEDIUM Confidence: Success Patterns\n\nOutput was accepted or praised.\n\n**Detection patterns:**\n- Explicit praise: \"perfect\", \"great\", \"yes\", \"exactly\"\n- User built on output without modification\n- Output was committed without changes\n\n**Example:**\n```text\nUser: \"Perfect, that's exactly what I needed\"\n→ [MED] + Add preference: \"Include usage examples in documentation\"\n```\n\n#### MEDIUM Confidence: Edge Cases\n\nScenarios not anticipated.\n\n**Detection patterns:**\n- Questions not answered\n- Workarounds user had to apply\n- Error handling gaps discovered\n\n#### LOW Confidence: Preferences\n\nAccumulated patterns over time.\n\n---\n\n### Phase 3: Propose Learnings\n\nPresent findings:\n\n```text\n┌─────────────────────────────────────────────────────────────┐\n│ REFLECTION: {target (agent\u002Fdecision\u002Fskill)}                  │\n├─────────────────────────────────────────────────────────────┤\n│                                                             │\n│ [HIGH] + Add constraint: \"{specific constraint}\"            │\n│   Source: \"{quoted user correction}\"                        │\n│   Target: tracked governing source via focused PR            │\n│                                                             │\n│ [MED]  + Add preference: \"{specific preference}\"            │\n│   Source: \"{evidence from conversation}\"                    │\n│   Target: governed runtime memory                            │\n│                                                             │\n│ [LOW]  ~ Note for review: \"{observation}\"                   │\n│   Source: \"{pattern observed}\"                              │\n│   Target: report only; no persistence                        │\n│                                                             │\n├─────────────────────────────────────────────────────────────┤\n│ Apply changes? [Y\u002Fn\u002Fedit]                                   │\n└─────────────────────────────────────────────────────────────┘\n```\n\n**Confidence Threshold:**\n\n| Threshold | Action |\n|-----------|--------|\n| ≥1 HIGH signal | Always propose (user explicitly corrected) |\n| ≥2 MED signals | Propose (sufficient pattern) |\n| ≥3 LOW signals | Propose (accumulated evidence) |\n| 1-2 LOW only | Skip (insufficient evidence) |\n\n### Phase 4: Persist Learnings\n\n**ALWAYS show changes before applying.**\n\nAfter user approval:\n\n1. **For Runtime Memory:**\n   - Use the configured governed memory or state tool.\n   - If no runtime persistence tool is available, report the learning without claiming it was saved.\n\n2. **For Behavioral Changes:**\n   - Route the accountable specialist to update the minimum tracked governing source.\n   - Validate the change and open a focused PR.\n\n3. **Never Use Throwaway Proposal Files:**\n   - Do not create local or gitignored Markdown proposals, including `.squad\u002Fdecisions\u002Finbox\u002F*.md`, for reflection or directive capture.\n   - Runtime memory may preserve context, but it does not replace the tracked source change required to alter behavior.\n\n---\n\n## Usage Examples\n\n### Example 1: User Correction\n\n**Conversation:**\n```\nAgent: \"I'll use grep to search the repository\"\nUser: \"No, use the code search tools first, grep is too slow\"\n```\n\n**Reflection Output:**\n```\n[HIGH] + Add constraint: \"Use code intelligence tools before grep\"\n  Source: \"No, use the code search tools first, grep is too slow\"\n  Target: accountable tracked instruction via focused PR; governed runtime\n          memory for context only\n```\n\n### Example 2: Success Pattern\n\n**Conversation:**\n```\nAgent: [Creates PR with detailed description and test plan]\nUser: \"Perfect! This is exactly the format I want for all PRs\"\n```\n\n**Reflection Output:**\n```\n[MED] + Add preference: \"Include test plan in PR descriptions\"\n  Source: User praised detailed PR format\n  Target: governed runtime memory; if adopted as team behavior, update the\n          accountable tracked instruction in a focused PR\n```\n\n---\n\n## When to Use\n\n✅ **Use reflect when:**\n- User says \"no\", \"wrong\", \"not like that\" (HIGH priority)\n- User says \"perfect\", \"exactly\", \"great\" (MED priority)\n- You discover edge cases or gaps\n- Complex work session with multiple learnings\n- At end of sprint\u002Fmilestone to consolidate patterns\n\n❌ **Don't use reflect when:**\n- Simple one-off questions with no pattern\n- User is just exploring ideas (no concrete decisions)\n- Learning is already captured in history.md\u002Fdecisions.md\n\n---\n\n## See Also\n\n- `.squad\u002Fdecisions.md` — Runtime-backed team decisions\n- `.github\u002Fagents\u002Fsquad.agent.md` — Coordinator behavior\n- `.squad\u002Frouting.md` — Work assignment patterns\n","# Reflect Skill\n\n**Critical learning capture system** for Squad. Prevents repeating mistakes and preserves successful patterns across sessions.\n\nAnalyze conversations and propose improvements to squad knowledge based on what worked, what didn't, and edge cases discovered. **Every correction is a learning opportunity.**\n\n---\n\n## Integration with Squad Architecture\n\n**Reflect complements existing Squad knowledge systems:**\n\n1. **Governed runtime memory** — Durable observations and decisions persisted through configured state tools\n2. **Tracked governing source** — Coordinator prompts, skills, and other behavior-defining files changed through focused pull requests\n3. **`reflect` skill** — Identifies learnings that may warrant either runtime memory or an approved tracked behavior change\n\n**Workflow:**\n- Use `reflect` during work to capture learnings\n- At session end, review captured learnings\n- Persist evidence through governed runtime memory when available\n- Promote accepted behavior changes through the accountable specialist as tracked source changes in a focused PR\n- Report adjacent improvements separately; do not expand the current scope without explicit user approval\n\n---\n\n## Triggers\n\n### ? HIGH Priority (Invoke Immediately)\n\n| Trigger | Example | Why Critical |\n|---------|---------|--------------|\n| User correction | \"no\", \"wrong\", \"not like that\", \"never do\" | Captures mistakes to prevent repetition |\n| Architectural insight | \"you removed that without understanding why\" | Documents design decisions (Chesterton's Fence) |\n| Immediate fixes | \"debug\", \"root cause\", \"fix all\" | Learns from errors in real-time |\n\n### ? MEDIUM Priority (Invoke After Multiple)\n\n| Trigger | Example | Why Important |\n|---------|---------|---------------|\n| User praise | \"perfect\", \"exactly\", \"great\" | Reinforces successful patterns |\n| Tool preferences | \"use X instead of Y\", \"prefer\" | Builds workflow preferences |\n| Edge cases | \"what if X happens?\", \"don't forget\", \"ensure\" | Captures scenarios to handle |\n\n### ? LOW Priority (Invoke at Session End)\n\n| Trigger | Example | Why Useful |\n|---------|---------|------------|\n| Repeated patterns | Frequent use of specific commands\u002Ftools | Identifies workflow preferences |\n| Session end | After complex work | Consolidates all session learnings |\n\n---\n\n## Process\n\n### Phase 1: Identify Learning Target\n\nDetermine what knowledge system should be updated:\n\n1. **Observation or preference** → Governed runtime memory, when configured\n2. **Accepted team-wide behavior change** → The accountable tracked instruction or skill source, through a focused PR\n3. **Adjacent improvement** → Report separately and require explicit user approval before adding it to scope\n\n### Phase 2: Analyze Conversation\n\nScan for learning signals with confidence levels:\n\n#### HIGH Confidence: Corrections\n\nUser actively steered or corrected output.\n\n**Detection patterns:**\n- Explicit rejection: \"no\", \"not like that\", \"that's wrong\"\n- Strong directives: \"never do\", \"always do\", \"don't ever\"\n- User provided alternative implementation\n\n**Example:**\n```text\nUser: \"No, use the azure-devops MCP tool instead of raw API calls\"\n→ [HIGH] + Add constraint: \"Prefer azure-devops MCP tools over REST API\"\n```\n\n#### MEDIUM Confidence: Success Patterns\n\nOutput was accepted or praised.\n\n**Detection patterns:**\n- Explicit praise: \"perfect\", \"great\", \"yes\", \"exactly\"\n- User built on output without modification\n- Output was committed without changes\n\n**Example:**\n```text\nUser: \"Perfect, that's exactly what I needed\"\n→ [MED] + Add preference: \"Include usage examples in documentation\"\n```\n\n#### MEDIUM Confidence: Edge Cases\n\nScenarios not anticipated.\n\n**Detection patterns:**\n- Questions not answered\n- Workarounds user had to apply\n- Error handling gaps discovered\n\n#### LOW Confidence: Preferences\n\nAccumulated patterns over time.\n\n---\n\n### Phase 3: Propose Learnings\n\nPresent findings:\n\n```text\n┌─────────────────────────────────────────────────────────────┐\n│ REFLECTION: {target (agent\u002Fdecision\u002Fskill)}                  │\n├─────────────────────────────────────────────────────────────┤\n│                                                             │\n│ [HIGH] + Add constraint: \"{specific constraint}\"            │\n│   Source: \"{quoted user correction}\"                        │\n│   Target: tracked governing source via focused PR            │\n│                                                             │\n│ [MED]  + Add preference: \"{specific preference}\"            │\n│   Source: \"{evidence from conversation}\"                    │\n│   Target: governed runtime memory                            │\n│                                                             │\n│ [LOW]  ~ Note for review: \"{observation}\"                   │\n│   Source: \"{pattern observed}\"                              │\n│   Target: report only; no persistence                        │\n│                                                             │\n├─────────────────────────────────────────────────────────────┤\n│ Apply changes? [Y\u002Fn\u002Fedit]                                   │\n└─────────────────────────────────────────────────────────────┘\n```\n\n**Confidence Threshold:**\n\n| Threshold | Action |\n|-----------|--------|\n| ≥1 HIGH signal | Always propose (user explicitly corrected) |\n| ≥2 MED signals | Propose (sufficient pattern) |\n| ≥3 LOW signals | Propose (accumulated evidence) |\n| 1-2 LOW only | Skip (insufficient evidence) |\n\n### Phase 4: Persist Learnings\n\n**ALWAYS show changes before applying.**\n\nAfter user approval:\n\n1. **For Runtime Memory:**\n   - Use the configured governed memory or state tool.\n   - If no runtime persistence tool is available, report the learning without claiming it was saved.\n\n2. **For Behavioral Changes:**\n   - Route the accountable specialist to update the minimum tracked governing source.\n   - Validate the change and open a focused PR.\n\n3. **Never Use Throwaway Proposal Files:**\n   - Do not create local or gitignored Markdown proposals, including `.squad\u002Fdecisions\u002Finbox\u002F*.md`, for reflection or directive capture.\n   - Runtime memory may preserve context, but it does not replace the tracked source change required to alter behavior.\n\n---\n\n## Usage Examples\n\n### Example 1: User Correction\n\n**Conversation:**\n```\nAgent: \"I'll use grep to search the repository\"\nUser: \"No, use the code search tools first, grep is too slow\"\n```\n\n**Reflection Output:**\n```\n[HIGH] + Add constraint: \"Use code intelligence tools before grep\"\n  Source: \"No, use the code search tools first, grep is too slow\"\n  Target: accountable tracked instruction via focused PR; governed runtime\n          memory for context only\n```\n\n### Example 2: Success Pattern\n\n**Conversation:**\n```\nAgent: [Creates PR with detailed description and test plan]\nUser: \"Perfect! This is exactly the format I want for all PRs\"\n```\n\n**Reflection Output:**\n```\n[MED] + Add preference: \"Include test plan in PR descriptions\"\n  Source: User praised detailed PR format\n  Target: governed runtime memory; if adopted as team behavior, update the\n          accountable tracked instruction in a focused PR\n```\n\n---\n\n## When to Use\n\n✅ **Use reflect when:**\n- User says \"no\", \"wrong\", \"not like that\" (HIGH priority)\n- User says \"perfect\", \"exactly\", \"great\" (MED priority)\n- You discover edge cases or gaps\n- Complex work session with multiple learnings\n- At end of sprint\u002Fmilestone to consolidate patterns\n\n❌ **Don't use reflect when:**\n- Simple one-off questions with no pattern\n- User is just exploring ideas (no concrete decisions)\n- Learning is already captured in history.md\u002Fdecisions.md\n\n---\n\n## See Also\n\n- `.squad\u002Fdecisions.md` — Runtime-backed team decisions\n- `.github\u002Fagents\u002Fsquad.agent.md` — Coordinator behavior\n- `.squad\u002Frouting.md` — Work assignment patterns\n",{"name":5,"domain":47,"license":15,"version":48,"confidence":28,"description":7},"team-memory, learning","1.0.0-squad","dev","7b731ec97f35346bdc4fa87c3edaf16b808fc0ae","templates\u002Fskills\u002Freflect",520,"https:\u002F\u002Fgithub.com\u002Fgithub\u002Fsquad","https:\u002F\u002Fgithub.github.io\u002Fsquad\u002F","https:\u002F\u002Fgithub.com\u002Fgithub","https:\u002F\u002Fgithub.com\u002Fabout","How people build software.","2026-11-15 02:13:59",{"source":8,"skillName":5,"commands":60,"prompt":70,"zipAvailable":71},[61,64,67],{"tool":62,"command":63},"npx","npx skills add github\u002Fsquad --skill reflect",{"tool":65,"command":66},"bunx","bunx skills add github\u002Fsquad --skill reflect",{"tool":68,"command":69},"pnpm","pnpm dlx skills add github\u002Fsquad --skill reflect","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fgithub\u002Fsquad\u002Fblob\u002Fdev\u002Ftemplates\u002Fskills\u002Freflect\u002FSKILL.md",true,[],[74,82,89,96,103,109],{"skillNo":75,"name":76,"nameZh":6,"description":77,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":78,"categories":79,"keywords":80,"publishedAt":81},"01M76752CVNE7K6NWD59CRAYNJ","windows-compatibility","Cross-platform path handling and command patterns","https:\u002F\u002Fgithub.com\u002Fgithub\u002Fsquad\u002Fblob\u002Fdev\u002Fplugins\u002Fsquad\u002Fskills\u002Fwindows-compatibility\u002FSKILL.md",[],[],"2026-11-12 13:55:43",{"skillNo":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":81},"01M767508W8A4G83NX1FKEW5VN","versioning-policy","SemVer rules for Squad's stable, preview, insider, and local package versions","https:\u002F\u002Fgithub.com\u002Fgithub\u002Fsquad\u002Fblob\u002Fdev\u002Fplugins\u002Fsquad\u002Fskills\u002Fversioning-policy\u002FSKILL.md",[],[],{"skillNo":90,"name":91,"nameZh":6,"description":92,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":93,"categories":94,"keywords":95,"publishedAt":81},"01M7674X5JJKNJ74B3SS8PGXG2","tiered-memory","Three-tier agent memory model (hot\u002Fcold\u002Fwiki) for context reduction per spawn","https:\u002F\u002Fgithub.com\u002Fgithub\u002Fsquad\u002Fblob\u002Fdev\u002Fplugins\u002Fsquad\u002Fskills\u002Ftiered-memory\u002FSKILL.md",[],[],{"skillNo":97,"name":98,"nameZh":6,"description":99,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":100,"categories":101,"keywords":102,"publishedAt":81},"01M7674V3S9H03A76H75M1KSAF","test-discipline","Design assertions that prove the actual contract and fail on realistic regressions","https:\u002F\u002Fgithub.com\u002Fgithub\u002Fsquad\u002Fblob\u002Fdev\u002Fplugins\u002Fsquad\u002Fskills\u002Ftest-discipline\u002FSKILL.md",[],[],{"skillNo":104,"name":41,"nameZh":6,"description":105,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":106,"categories":107,"keywords":108,"publishedAt":81},"01M7674QB8CA61RNDX2Z0NAK71","Squad's command catalog and interactive menu. Invoke via \u002Fsquad (slash command) or natural language (\"squad commands\", \"what can squad do\", \"show me squad options\"). Presents categorized operations (Install & Upgrade, Team Management, Issues & PRs, Plugins & Skills, Model & Cost, Sessions & State) as an interactive picker. Routes to the right squad CLI command or the Squad coordinator agent.","https:\u002F\u002Fgithub.com\u002Fgithub\u002Fsquad\u002Fblob\u002Fdev\u002Fplugins\u002Fsquad\u002Fskills\u002Fsquad\u002FSKILL.md",[],[],{"skillNo":110,"name":111,"nameZh":6,"description":112,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":113,"categories":114,"keywords":115,"publishedAt":81},"01M7674KTQ632N30HQB1TP7TXH","squad-version-check","Internals of how @bradygaster\u002Fsquad-cli stamps its version, how `squad upgrade` works (what it preserves vs overwrites), and how to probe the npm registry for the latest version from a coordinator prompt.","https:\u002F\u002Fgithub.com\u002Fgithub\u002Fsquad\u002Fblob\u002Fdev\u002Fplugins\u002Fsquad\u002Fskills\u002Fsquad-version-check\u002FSKILL.md",[],[]]