[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M75WDQRSZREJF4CSTMC7PP0Q":3,"skill-same:01M75WDQRSZREJF4CSTMC7PP0Q":45},{"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":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":25,"commitSha":26,"skillDirectory":27,"forksCount":14,"repositoryHtmlUrl":28,"repositoryHomepageUrl":6,"ownerHtmlUrl":29,"ownerBlogUrl":6,"ownerDescription":30,"updatedAt":31,"install":32,"resources":44},"01M75WDQRSZREJF4CSTMC7PP0Q","plan",null,"Create a detailed implementation plan","stevengonsalvez\u002Fainb-toolkit","stevengonsalvez","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F9320602?v=4","01M75W59QRYZYANWY3XG8CJB1P","Steven Gonsalvez","01M75W59RD5YW1RT67BDV0EAZF",1,"MIT","https:\u002F\u002Fgithub.com\u002Fstevengonsalvez\u002Fainb-toolkit\u002Fblob\u002Fmain\u002Fskills\u002Fplan\u002FSKILL.md",[],[],"2026-11-12 13:55:43",false,"---\nname: plan\ndescription: Create a detailed implementation plan\nuser-invocable: true\n---\n\n# Create Plan\n\nProduce a phased, file-level implementation plan through interactive research and iteration. Be skeptical: verify every requirement against actual code before writing it into the plan. Output is a `plans\u002F*.md` file that `\u002Fimplement` consumes directly.\n\n**When NOT to use \u002F route elsewhere:**\n- Requirements still fuzzy, no clear task yet → run `\u002Finterview` first, then return here.\n- No codebase understanding yet, need to explore what exists → run `\u002Fresearch` first (it searches learnings + codebase); its output feeds Step 1 here.\n- Test-first workflow wanted → use `\u002Fplan-tdd` instead.\n\n\u003C!-- recall:begin -->\n\n## Step 0: Prior-art check (MANDATORY, run first)\n\nRecall prior learnings so you don't re-decide something already captured:\n\n```bash\nuv run \"{{HOME_TOOL_DIR}}\u002Fskills\u002Frecall\u002Fscripts\u002Frecall.py\" \\\n  \"\u003CQUERY>\" \\\n  --limit 5 --format markdown\n```\n\n- `\u003CQUERY>` = user task description + any file paths + domain keywords (e.g. `\"user auth OAuth migration\"`).\n- Result names a constraint \u002F anti-pattern \u002F prior decision relevant to the task → surface it to the user BEFORE the main flow.\n- Empty output or non-zero exit → treat as \"no prior art\", proceed silently. Never block on recall failure.\n\n\u003C!-- recall:end -->\n\n## Step 1: Intake\n\n1. Parameters (file path or task description) provided → skip the greeting, read provided files FULLY (Read tool, no `limit`\u002F`offset`), go to Step 2.\n2. No parameters → print exactly this, then wait:\n\n```\nI'll help you create a detailed implementation plan. Let me start by understanding what we're building.\n\nPlease provide:\n1. The task\u002Frequirement description\n2. Any relevant context, constraints, or specific requirements\n3. Links to related research or previous implementations\n\nI can also check for existing research documents if you've already run \u002Fresearch on this topic.\n```\n\n## Step 2: Context gathering\n\nDo these in order. Read every file FULLY before spawning any sub-task.\n\n| Source | Action |\n|--------|--------|\n| `research\u002F` dir | Relevant file exists → read it as the plan's foundation. None → run `\u002Fresearch`, then continue. |\n| `.planning\u002FROADMAP.md` | Exists → read; use its phases as the plan's phase structure. |\n| `.planning\u002FSTATE.md` | Exists → read current position + blockers. |\n| `.planning\u002F**\u002FCONTEXT.md` | Exists → read locked decisions; **the plan MUST honor these**. |\n| Any file the user mentioned | Read FULLY yourself before delegating. |\n\nThen spawn **general-purpose** sub-agents in parallel (one message, multiple Task calls) to: (a) find all files related to the task, (b) map the current implementation, (c) find existing docs. Each returns explanations with `file:line` references. Wait for all, then read every file they identified FULLY.\n\nCross-reference requirements against the real code. Note discrepancies and assumptions needing verification.\n\nPresent understanding + only the questions research could not answer:\n\n```\nBased on my research of the codebase, I understand we need to [accurate summary].\n\nI've found that:\n- [Current implementation detail with file:line reference]\n- [Relevant pattern or constraint discovered]\n- [Potential complexity or edge case identified]\n\nQuestions that my research couldn't answer:\n- [Specific technical question requiring human judgment]\n- [Business logic clarification]\n```\n\n## Step 3: Deep research & design options\n\n- User corrects a misunderstanding → do NOT just accept it. Spawn a new research task \u002F read the named files, verify the fact yourself, THEN proceed.\n- Track research tasks with TodoWrite.\n- Spawn parallel sub-tasks for: deeper file discovery, patterns\u002Fconventions to follow, integration points, similar features to model after, existing tests. Wait for ALL.\n- Present findings + design options and get the user to pick:\n\n```\nBased on my research, here's what I found:\n\n**Current State:**\n- [Key discovery about existing code]\n\n**Design Options:**\n1. [Option A] - [pros\u002Fcons]\n2. [Option B] - [pros\u002Fcons]\n\nWhich approach aligns best with your vision?\n```\n\n## Step 4: Phase structure + wave dependencies\n\n1. Propose phases (name + what each accomplishes). Confirm phasing with the user before writing details.\n\n2. Assign waves. **Wave** = a set of phases that can run in parallel.\n\n| Rule | Meaning |\n|------|---------|\n| No dependencies | Wave 1 |\n| Depends on a Wave N phase | Wave N+1 |\n| File appears in 2 phases of the SAME wave | NOT allowed — add an artificial dependency to serialize them |\n\nPresent the dependency graph, e.g.:\n\n```\nPhase 1: No dependencies (Wave 1)\nPhase 2: Depends on Phase 1 (Wave 2)\nPhase 3: No dependencies (Wave 1) -- parallel with Phase 1\nPhase 4: Depends on Phase 2, Phase 3 (Wave 3)\n```\n\nGet sign-off on structure before writing the full plan.\n\n## Step 5: Write the plan\n\nWrite to `plans\u002F{descriptive_name}.md`. Use this exact skeleton — `\u002Fimplement` parses the `\u003C!-- wave: ... -->` comments and `[CHECKPOINT:*]` markers, so keep both formats verbatim:\n\n```markdown\n# [Feature\u002FTask Name] Implementation Plan\n\n## Overview\n[What we're implementing and why, 1-2 sentences]\n\n## Current State Analysis\n[What exists now, what's missing, key constraints discovered]\n\n## Desired End State\n[The end state and how to verify it]\n\n### Key Discoveries:\n- [Finding with file:line reference]\n- [Pattern to follow]\n- [Constraint to work within]\n\n## What We're NOT Doing\n[Explicit out-of-scope items — prevents scope creep]\n\n## Implementation Approach\n[High-level strategy and reasoning]\n\n## Phase 1: [Descriptive Name]\n\u003C!-- wave: 1 | depends_on: [] | files: [path\u002Fto\u002Ffile1.ext, path\u002Fto\u002Ffile2.ext] -->\n\n### Overview\n[What this phase accomplishes]\n\n### Changes Required:\n\n#### 1. [Component\u002FFile Group]\n**File**: `path\u002Fto\u002Ffile.ext`\n**Changes**: [Summary]\n\n```[language]\n\u002F\u002F Specific code to add\u002Fmodify\n```\n\n### Success Criteria:\n\n#### Automated Verification:\n- [ ] Tests pass: `npm test`\n- [ ] Type checking passes: `npm run typecheck`\n- [ ] Linting passes: `npm run lint`\n- [ ] Build succeeds: `npm run build`\n\n#### Manual Verification:\n- [ ] Feature works as expected when tested\n- [ ] Edge cases handled correctly\n- [ ] No regressions in related features\n\n### Checkpoints (if applicable):\n- **`[CHECKPOINT:human-verify]`**: Review automated work before continuing\n  - What was built: [description]\n  - How to verify: [numbered steps + expected outcomes]\n  - Resume: Type \"approved\" or describe issues\n- **`[CHECKPOINT:decision]`**: Choose between options\n  - Options: [A vs B with trade-offs]\n  - Impact: [what changes based on choice]\n- **`[CHECKPOINT:human-action]`**: Non-automatable step (rare)\n  - What's needed: [action only a human can take, e.g. \"Click email verification link\"]\n\n---\n\n## Phase 2: [Descriptive Name]\n[Same structure]\n\n---\n\n## Testing Strategy\n### Unit Tests: [what to test, key edge cases]\n### Integration Tests: [end-to-end scenarios]\n### Manual Testing Steps:\n1. [Step to verify feature]\n2. [Edge case to test manually]\n\n## Performance Considerations\n[Implications or optimizations, if any]\n\n## Migration Notes\n[How to handle existing data\u002Fsystems, if applicable]\n\n## References\n- Original requirements: [location]\n- Related research: `research\u002F[relevant].md`\n- Similar implementation: `[file:line]`\n```\n\n**Hard rules for the written plan:**\n\n| Rule | Detail |\n|------|--------|\n| File ownership | Each file may be modified in only ONE phase per wave. Overlap within a wave → add a dependency between those phases. List every file a phase touches in its `files:` wave comment. |\n| Checkpoint: automatable | If Claude CAN do it via CLI\u002FAPI\u002FBash, it MUST NOT be a checkpoint. |\n| Checkpoint: budget | Max 1 checkpoint per phase (prevents fatigue). |\n| Checkpoint: mix | `human-verify` ≈ 90% (visual\u002FUX review); `decision` occasional; `human-action` ≈ 1% (only what Claude literally cannot do). |\n| Success criteria | Always split into Automated (runnable commands, file existence, compile\u002Ftypecheck\u002Ftests) vs Manual (UI\u002FUX, real-world perf, hard-to-automate edge cases). |\n| No open questions | Every decision resolved before finalizing. Research or ask immediately — never leave unresolved questions in the final plan. |\n\n## Step 6: Review & iterate\n\nTell the user the plan location and ask for review:\n\n```\nI've created the initial implementation plan at:\n`plans\u002F[filename].md`\n\nPlease review it and let me know:\n- Are the phases properly scoped?\n- Are the success criteria specific enough?\n- Any technical details that need adjustment?\n- Missing edge cases or considerations?\n```\n\nIterate on feedback (add\u002Fremove phases, adjust approach, sharpen criteria, adjust scope) until the user is satisfied.\n\n## Phase-ordering patterns (pick by task type)\n\n| Task type | Phase order |\n|-----------|-------------|\n| New feature | Research existing patterns → data model → backend logic → API endpoints → UI last |\n| Refactoring | Document current behavior → incremental changes → maintain backwards compat → migration strategy |\n| Database change | Schema\u002Fmigration → data access methods → business logic → API → clients |\n","# Create Plan\n\nProduce a phased, file-level implementation plan through interactive research and iteration. Be skeptical: verify every requirement against actual code before writing it into the plan. Output is a `plans\u002F*.md` file that `\u002Fimplement` consumes directly.\n\n**When NOT to use \u002F route elsewhere:**\n- Requirements still fuzzy, no clear task yet → run `\u002Finterview` first, then return here.\n- No codebase understanding yet, need to explore what exists → run `\u002Fresearch` first (it searches learnings + codebase); its output feeds Step 1 here.\n- Test-first workflow wanted → use `\u002Fplan-tdd` instead.\n\n\u003C!-- recall:begin -->\n\n## Step 0: Prior-art check (MANDATORY, run first)\n\nRecall prior learnings so you don't re-decide something already captured:\n\n```bash\nuv run \"{{HOME_TOOL_DIR}}\u002Fskills\u002Frecall\u002Fscripts\u002Frecall.py\" \\\n  \"\u003CQUERY>\" \\\n  --limit 5 --format markdown\n```\n\n- `\u003CQUERY>` = user task description + any file paths + domain keywords (e.g. `\"user auth OAuth migration\"`).\n- Result names a constraint \u002F anti-pattern \u002F prior decision relevant to the task → surface it to the user BEFORE the main flow.\n- Empty output or non-zero exit → treat as \"no prior art\", proceed silently. Never block on recall failure.\n\n\u003C!-- recall:end -->\n\n## Step 1: Intake\n\n1. Parameters (file path or task description) provided → skip the greeting, read provided files FULLY (Read tool, no `limit`\u002F`offset`), go to Step 2.\n2. No parameters → print exactly this, then wait:\n\n```\nI'll help you create a detailed implementation plan. Let me start by understanding what we're building.\n\nPlease provide:\n1. The task\u002Frequirement description\n2. Any relevant context, constraints, or specific requirements\n3. Links to related research or previous implementations\n\nI can also check for existing research documents if you've already run \u002Fresearch on this topic.\n```\n\n## Step 2: Context gathering\n\nDo these in order. Read every file FULLY before spawning any sub-task.\n\n| Source | Action |\n|--------|--------|\n| `research\u002F` dir | Relevant file exists → read it as the plan's foundation. None → run `\u002Fresearch`, then continue. |\n| `.planning\u002FROADMAP.md` | Exists → read; use its phases as the plan's phase structure. |\n| `.planning\u002FSTATE.md` | Exists → read current position + blockers. |\n| `.planning\u002F**\u002FCONTEXT.md` | Exists → read locked decisions; **the plan MUST honor these**. |\n| Any file the user mentioned | Read FULLY yourself before delegating. |\n\nThen spawn **general-purpose** sub-agents in parallel (one message, multiple Task calls) to: (a) find all files related to the task, (b) map the current implementation, (c) find existing docs. Each returns explanations with `file:line` references. Wait for all, then read every file they identified FULLY.\n\nCross-reference requirements against the real code. Note discrepancies and assumptions needing verification.\n\nPresent understanding + only the questions research could not answer:\n\n```\nBased on my research of the codebase, I understand we need to [accurate summary].\n\nI've found that:\n- [Current implementation detail with file:line reference]\n- [Relevant pattern or constraint discovered]\n- [Potential complexity or edge case identified]\n\nQuestions that my research couldn't answer:\n- [Specific technical question requiring human judgment]\n- [Business logic clarification]\n```\n\n## Step 3: Deep research & design options\n\n- User corrects a misunderstanding → do NOT just accept it. Spawn a new research task \u002F read the named files, verify the fact yourself, THEN proceed.\n- Track research tasks with TodoWrite.\n- Spawn parallel sub-tasks for: deeper file discovery, patterns\u002Fconventions to follow, integration points, similar features to model after, existing tests. Wait for ALL.\n- Present findings + design options and get the user to pick:\n\n```\nBased on my research, here's what I found:\n\n**Current State:**\n- [Key discovery about existing code]\n\n**Design Options:**\n1. [Option A] - [pros\u002Fcons]\n2. [Option B] - [pros\u002Fcons]\n\nWhich approach aligns best with your vision?\n```\n\n## Step 4: Phase structure + wave dependencies\n\n1. Propose phases (name + what each accomplishes). Confirm phasing with the user before writing details.\n\n2. Assign waves. **Wave** = a set of phases that can run in parallel.\n\n| Rule | Meaning |\n|------|---------|\n| No dependencies | Wave 1 |\n| Depends on a Wave N phase | Wave N+1 |\n| File appears in 2 phases of the SAME wave | NOT allowed — add an artificial dependency to serialize them |\n\nPresent the dependency graph, e.g.:\n\n```\nPhase 1: No dependencies (Wave 1)\nPhase 2: Depends on Phase 1 (Wave 2)\nPhase 3: No dependencies (Wave 1) -- parallel with Phase 1\nPhase 4: Depends on Phase 2, Phase 3 (Wave 3)\n```\n\nGet sign-off on structure before writing the full plan.\n\n## Step 5: Write the plan\n\nWrite to `plans\u002F{descriptive_name}.md`. Use this exact skeleton — `\u002Fimplement` parses the `\u003C!-- wave: ... -->` comments and `[CHECKPOINT:*]` markers, so keep both formats verbatim:\n\n```markdown\n# [Feature\u002FTask Name] Implementation Plan\n\n## Overview\n[What we're implementing and why, 1-2 sentences]\n\n## Current State Analysis\n[What exists now, what's missing, key constraints discovered]\n\n## Desired End State\n[The end state and how to verify it]\n\n### Key Discoveries:\n- [Finding with file:line reference]\n- [Pattern to follow]\n- [Constraint to work within]\n\n## What We're NOT Doing\n[Explicit out-of-scope items — prevents scope creep]\n\n## Implementation Approach\n[High-level strategy and reasoning]\n\n## Phase 1: [Descriptive Name]\n\u003C!-- wave: 1 | depends_on: [] | files: [path\u002Fto\u002Ffile1.ext, path\u002Fto\u002Ffile2.ext] -->\n\n### Overview\n[What this phase accomplishes]\n\n### Changes Required:\n\n#### 1. [Component\u002FFile Group]\n**File**: `path\u002Fto\u002Ffile.ext`\n**Changes**: [Summary]\n\n```[language]\n\u002F\u002F Specific code to add\u002Fmodify\n```\n\n### Success Criteria:\n\n#### Automated Verification:\n- [ ] Tests pass: `npm test`\n- [ ] Type checking passes: `npm run typecheck`\n- [ ] Linting passes: `npm run lint`\n- [ ] Build succeeds: `npm run build`\n\n#### Manual Verification:\n- [ ] Feature works as expected when tested\n- [ ] Edge cases handled correctly\n- [ ] No regressions in related features\n\n### Checkpoints (if applicable):\n- **`[CHECKPOINT:human-verify]`**: Review automated work before continuing\n  - What was built: [description]\n  - How to verify: [numbered steps + expected outcomes]\n  - Resume: Type \"approved\" or describe issues\n- **`[CHECKPOINT:decision]`**: Choose between options\n  - Options: [A vs B with trade-offs]\n  - Impact: [what changes based on choice]\n- **`[CHECKPOINT:human-action]`**: Non-automatable step (rare)\n  - What's needed: [action only a human can take, e.g. \"Click email verification link\"]\n\n---\n\n## Phase 2: [Descriptive Name]\n[Same structure]\n\n---\n\n## Testing Strategy\n### Unit Tests: [what to test, key edge cases]\n### Integration Tests: [end-to-end scenarios]\n### Manual Testing Steps:\n1. [Step to verify feature]\n2. [Edge case to test manually]\n\n## Performance Considerations\n[Implications or optimizations, if any]\n\n## Migration Notes\n[How to handle existing data\u002Fsystems, if applicable]\n\n## References\n- Original requirements: [location]\n- Related research: `research\u002F[relevant].md`\n- Similar implementation: `[file:line]`\n```\n\n**Hard rules for the written plan:**\n\n| Rule | Detail |\n|------|--------|\n| File ownership | Each file may be modified in only ONE phase per wave. Overlap within a wave → add a dependency between those phases. List every file a phase touches in its `files:` wave comment. |\n| Checkpoint: automatable | If Claude CAN do it via CLI\u002FAPI\u002FBash, it MUST NOT be a checkpoint. |\n| Checkpoint: budget | Max 1 checkpoint per phase (prevents fatigue). |\n| Checkpoint: mix | `human-verify` ≈ 90% (visual\u002FUX review); `decision` occasional; `human-action` ≈ 1% (only what Claude literally cannot do). |\n| Success criteria | Always split into Automated (runnable commands, file existence, compile\u002Ftypecheck\u002Ftests) vs Manual (UI\u002FUX, real-world perf, hard-to-automate edge cases). |\n| No open questions | Every decision resolved before finalizing. Research or ask immediately — never leave unresolved questions in the final plan. |\n\n## Step 6: Review & iterate\n\nTell the user the plan location and ask for review:\n\n```\nI've created the initial implementation plan at:\n`plans\u002F[filename].md`\n\nPlease review it and let me know:\n- Are the phases properly scoped?\n- Are the success criteria specific enough?\n- Any technical details that need adjustment?\n- Missing edge cases or considerations?\n```\n\nIterate on feedback (add\u002Fremove phases, adjust approach, sharpen criteria, adjust scope) until the user is satisfied.\n\n## Phase-ordering patterns (pick by task type)\n\n| Task type | Phase order |\n|-----------|-------------|\n| New feature | Research existing patterns → data model → backend logic → API endpoints → UI last |\n| Refactoring | Document current behavior → incremental changes → maintain backwards compat → migration strategy |\n| Database change | Schema\u002Fmigration → data access methods → business logic → API → clients |\n",{"name":5,"description":7,"user-invocable":24},true,"main","9cf32d1e636db09530158af14ba301bfa5d401e9","skills\u002Fplan","https:\u002F\u002Fgithub.com\u002Fstevengonsalvez\u002Fainb-toolkit","https:\u002F\u002Fgithub.com\u002Fstevengonsalvez","stevengonsalvez.com","2026-11-12 00:50:59",{"source":8,"skillName":5,"commands":33,"prompt":43,"zipAvailable":24},[34,37,40],{"tool":35,"command":36},"npx","npx skills add stevengonsalvez\u002Fainb-toolkit --skill plan",{"tool":38,"command":39},"bunx","bunx skills add stevengonsalvez\u002Fainb-toolkit --skill plan",{"tool":41,"command":42},"pnpm","pnpm dlx skills add stevengonsalvez\u002Fainb-toolkit --skill plan","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fstevengonsalvez\u002Fainb-toolkit\u002Fblob\u002Fmain\u002Fskills\u002Fplan\u002FSKILL.md",[],[46,53,60,66,73,80],{"skillNo":47,"name":48,"nameZh":6,"description":49,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":50,"categories":51,"keywords":52,"publishedAt":19},"01M75WK9B8DKVJBJJ4CX2QMJMX","well-formed","Fixture skill that satisfies every mechanical rule, used to prove the conformance check passes a clean tree instead of failing everything.","https:\u002F\u002Fgithub.com\u002Fstevengonsalvez\u002Fainb-toolkit\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fagentskills\u002Fclean\u002Fwell-formed\u002FSKILL.md",[],[],{"skillNo":54,"name":55,"nameZh":6,"description":56,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":57,"categories":58,"keywords":59,"publishedAt":19},"01M75WK6ZGDXJJN8879K49QEYG","underscore-tools","Fixture skill that spells the tool allowlist key with an underscore instead of the hyphen the spec requires.","https:\u002F\u002Fgithub.com\u002Fstevengonsalvez\u002Fainb-toolkit\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fagentskills\u002Fbroken\u002Funderscore-tools\u002FSKILL.md",[],[],{"skillNo":61,"name":62,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":63,"categories":64,"keywords":65,"publishedAt":19},"01M75WK524MVT2H9E27ZCEA3HR","no-frontmatter","https:\u002F\u002Fgithub.com\u002Fstevengonsalvez\u002Fainb-toolkit\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fagentskills\u002Fbroken\u002Fno-frontmatter\u002FSKILL.md",[],[],{"skillNo":67,"name":68,"nameZh":6,"description":69,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":70,"categories":71,"keywords":72,"publishedAt":19},"01M75WJWZBMKR8R60PG3T3MCCC","long-description","Fixture skill with a deliberately oversized description used to prove the conformance check rejects anything past the 1024 character ceiling. Fixture skill with a deliberately oversized description used to prove the conformance check rejects anything past the 1024 character ceiling. Fixture skill with a deliberately oversized description used to prove the conformance check rejects anything past the 1024 character ceiling. Fixture skill with a deliberately oversized description used to prove the conformance check rejects anything past the 1024 character ceiling. Fixture skill with a deliberately oversized description used to prove the conformance check rejects anything past the 1024 character ceiling. Fixture skill with a deliberately oversized description used to prove the conformance check rejects anything past the 1024 character ceiling. Fixture skill with a deliberately oversized description used to prove the conformance check rejects anything past the 1024 character ceiling. Fixture skill with a deliberately oversized description used to prove the conformance check rejects anything past the 1024 character ceiling. Fixture skill with a deliberately oversized description used to prove the conformance check rejects anything past the 1024 character ceiling. Fixture skill with a deliberately oversized description used to prove the conformance check rejects anything past the 1024 character ceiling. Fixture skill with a deliberately oversized description used to prove the conformance check rejects anything past the 1024 character ceiling. Fixture skill with a deliberately oversized description used to prove the conformance check rejects anything past the 1024 character ceiling.","https:\u002F\u002Fgithub.com\u002Fstevengonsalvez\u002Fainb-toolkit\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fagentskills\u002Fbroken\u002Flong-description\u002FSKILL.md",[],[],{"skillNo":74,"name":75,"nameZh":6,"description":76,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":77,"categories":78,"keywords":79,"publishedAt":19},"01M75WJVRH3J22WKED7E5D01S8","some-other-name","Fixture skill whose frontmatter name does not match its parent directory name.","https:\u002F\u002Fgithub.com\u002Fstevengonsalvez\u002Fainb-toolkit\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fagentskills\u002Fbroken\u002Fdir-mismatch\u002FSKILL.md",[],[],{"skillNo":81,"name":82,"nameZh":6,"description":83,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":84,"categories":85,"keywords":86,"publishedAt":19},"01M75WJTC7H5FRP6S129VBSMZC","bad_Name","Fixture skill whose name uses an underscore and a capital letter, so it violates the agentskills.io name pattern.","https:\u002F\u002Fgithub.com\u002Fstevengonsalvez\u002Fainb-toolkit\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fagentskills\u002Fbroken\u002Fbad_Name\u002FSKILL.md",[],[]]