[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7C1PSQ896JSJ5VS3HR46XN8":3,"skill-same:01M7C1PSQ896JSJ5VS3HR46XN8":58},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":23,"commitSha":24,"skillDirectory":25,"forksCount":26,"repositoryHtmlUrl":27,"repositoryHomepageUrl":6,"ownerHtmlUrl":28,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":29,"install":30,"resources":43},"01M7C1PSQ896JSJ5VS3HR46XN8","implement-tasks",null,"Use when executing a batch of TDD-sized tasks inside a run-iteration call — dispatches an implementer subagent per task following red-green-refactor discipline and returns per-task completion status.","zak-keown\u002Fmoe","zak-keown","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F320717241?v=4","01M7BZFXME0NW6TQX0H7AZM2KS","01M7BZFXN32VR0JN5AATRD53H4",0,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002F.pi\u002Fskills\u002Fimplement-tasks\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: implement-tasks\ndescription: Use when executing a batch of TDD-sized tasks inside a run-iteration call — dispatches an implementer subagent per task following red-green-refactor discipline and returns per-task completion status.\n---\n\n# Implementing Tasks\n\n## Overview\n\nTakes an in-memory batch of TDD-sized tasks and executes each through: implementer subagent (TDD) → PAR spec-compliance review → fix loop → PAR code-quality review with boxing-in check → fix loop → mark complete. This is a fork of `sdd` — a sibling skill in this same plugin — with the plan-file reading phase stripped and the final end-of-plan reviewer removed. Reach for `sdd` when you have a written plan file to execute; this skill is only invoked by `run-iteration`, which holds its task batch in memory.\n\n## When to Use\n\nInvoked by `run-iteration` with a list of tasks. Tasks are passed in memory, not via a file.\n\n## Validate the batch\n\nBefore dispatching anything, validate every in-memory task has a non-empty\n`Files:` field, an `Interfaces:` field, and explicit `Consumes:` and `Produces:`\nentries. `None` is the explicit value when no interface edge exists. Missing\nany field fails batch validation and returns the batch to the caller; do not\ninfer metadata from prose or execute the malformed task sequentially.\n\nFor a valid batch, use exactly two execution rungs:\n\n1. worktree-isolated parallel dispatch when files are pairwise disjoint, there\n   is no in-wave `Consumes:` → `Produces:` edge, and every worker has a\n   pairwise-unique linked Git directory;\n2. sequential dispatch of the whole group when the gate does not hold or any\n   worktree cannot be created or validated.\n\nThere is no unisolated-parallel rung. Create and validate every worktree before\ndispatching any worker so a failed setup cannot produce a partially parallel\ngroup.\n\nBefore creating a parallel group, record the controller state:\n\n```bash\nCONTROLLER_BRANCH=$(git branch --show-current)\nCONTROLLER_HEAD=$(git rev-parse HEAD)\n```\n\nEvery worker in that group must branch from exactly `CONTROLLER_HEAD`. Record\neach owned worker path and branch before dispatch. A later group cannot start\nuntil the current group's reviewed commits have been integrated into the\ncontroller.\n\n## Per-Task Cycle\n\nFor each task in the provided list:\n\n### 1. Dispatch implementer\n\nUsing the template in `implementer-subagent-prompt.md`, dispatch a single implementer subagent with:\n- The full task description and context\n- The proof obligations for each observable AC in the task's stories\n- The list of existing scenarios that may be impacted\n\nThe implementer MUST complete a pre-flight mapping (AC → proof seam → scenario) before writing code. If the implementer skips the pre-flight, re-dispatch with explicit instructions to complete it first.\n\n### 2. Handle implementer status\n\n- **DONE:** proceed to spec-compliance review (step 3). Verify the implementer's report includes pre-flight mapping and scenario updates.\n- **DONE_WITH_CONCERNS:** read the concerns. If about correctness\u002Fscope, address before review. If observations, note and proceed.\n- **NEEDS_CONTEXT:** provide the missing context and re-dispatch\n- **BLOCKED:** assess: context problem → re-dispatch with context; too hard → re-dispatch with more capable model; task too large → break into smaller pieces; plan wrong → escalate to caller\n\n### 3. PAR spec-compliance review (Stage 1)\n\nFollowing [skills\u002F_shared\u002Fparallel-adversarial-review.md](..\u002F_shared\u002Fparallel-adversarial-review.md), resolved relative to this loaded document:\n\n1. Build spec-compliance prompt using `spec-compliance-reviewer-prompt.md`\n   - Include the proof obligations and the implementer's evidence claims\n2. Resolve [skills\u002F_shared\u002Fpar-reviewer-wrapper.md](..\u002F_shared\u002Fpar-reviewer-wrapper.md) relative to this loaded document and wrap in its competitive framing\n3. Dispatch TWO spec-compliance reviewers in parallel\n4. Aggregate findings (PAR rules: union of findings, severity = take worst)\n5. If ❌ issues found:\n   - Send aggregated issues back to the implementer subagent (same subagent, via continuation message)\n   - Implementer fixes\n   - Re-dispatch fresh PAR spec-compliance pair\n   - Repeat until ✅ spec compliant with adequate evidence\n6. Only proceed to Stage 2 after Stage 1 is ✅\n\n### 4. PAR code-quality review (Stage 2)\n\nFollowing [skills\u002F_shared\u002Fparallel-adversarial-review.md](..\u002F_shared\u002Fparallel-adversarial-review.md), resolved relative to this loaded document:\n\n1. Build code-quality prompt using `code-quality-reviewer-prompt.md`\n   - Include the next 3 pending roadmap iterations for the boxing-in check\n   - Include the implementer's corpus contribution for quality review\n2. Wrap in PAR competitive framing\n3. Dispatch TWO code-quality reviewers in parallel\n4. Aggregate findings\n5. If ❌ changes needed:\n   - Send aggregated issues back to the implementer\n   - Implementer fixes\n   - Re-dispatch fresh PAR code-quality pair\n   - Repeat until ✅ approved\n\n### 5. Mark task reviewed\n\nRecord the task as reviewed. In sequential mode it is also integrated and may\nbe marked complete immediately. In parallel mode it remains pending integration\nuntil every worker in its group passes both review stages.\n\n## Integrate the Parallel Wave\n\nAfter every worker in a parallel group is reviewed:\n\n1. Return to the controller workspace and verify its branch and HEAD still\n   equal `CONTROLLER_BRANCH` and `CONTROLLER_HEAD`. If either moved, stop; a\n   concurrent change invalidated the recorded merge target.\n2. Merge each worker branch into the controller, one at a time. Stop on the\n   first conflict and preserve every worktree for investigation.\n3. Run the iteration's complete test set on the integrated controller HEAD.\n   Per-task evidence from separate branches does not prove the combined tree.\n4. When the integrated tests pass, remove each owned worker worktree, prune\n   stale registrations, and delete its merged worker branch. Never remove a\n   host-owned worktree or force removal of an unclean worktree.\n5. Mark the group's tasks complete and record the new controller HEAD. Use that\n   integrated HEAD as the base for the next group.\n\nThe controller must contain every completed task before returning to\n`run-iteration`. Returning only per-task reports while commits remain on\nworker branches is a failed iteration, not successful parallel execution.\n\nAfter all tasks are integrated, return a per-task result list to the caller, including:\n- Per-task status\n- Scenarios added or updated per task\n- Evidence commands per task\n\n## Model Selection\n\nUse the least powerful model that can handle each role:\n\n| Role | Signal → Model |\n|---|---|\n| Implementer (mechanical: 1-2 files, clear spec) | Cheap\u002Ffast model |\n| Implementer (integration: multi-file, judgment) | Standard model |\n| Spec-compliance reviewer | Standard model |\n| Code-quality reviewer | Most capable model |\n\n## Quick Reference\n\n| Per task | Subagents dispatched |\n|---|---|\n| Implementer | 1 (sequential, TDD) |\n| Spec-compliance review (PAR) | 2 in parallel |\n| Code-quality review (PAR) | 2 in parallel |\n| **Minimum per task** | **5** (before re-review loops) |\n\n## Red Flags\n\n- **Never** start code-quality review before spec compliance is ✅\n- **Never** skip the re-review after fixes (reviewer found issues = implementer fixes = review again)\n- **Do not** dispatch multiple implementers concurrently unless the batch's\n  tasks pass rung one of the worktree gate: disjoint `Files:` blocks, no\n  `Consumes:` → `Produces:` edge inside the concurrent group, and a validated,\n  pairwise-unique linked Git directory per worker (see\n  `dispatch-agents` for the full gate and `use-worktrees`\n  Step 1d). File\u002Fdependency conflicts or any worktree setup failure select rung\n  two for the whole group: sequential dispatch. Never dispatch in parallel\n  without isolation.\n- **Never** accept \"close enough\" on spec compliance\n- **Never** let implementer self-review replace the two-stage review\n\n## References\n\n- `implementer-subagent-prompt.md` — implementer dispatch template\n- `spec-compliance-reviewer-prompt.md` — Stage 1 review template\n- `code-quality-reviewer-prompt.md` — Stage 2 review template (includes boxing-in)\n- [skills\u002F_shared\u002Fparallel-adversarial-review.md](..\u002F_shared\u002Fparallel-adversarial-review.md) — PAR methodology; resolve relative to this loaded document\n- [skills\u002F_shared\u002Fpar-reviewer-wrapper.md](..\u002F_shared\u002Fpar-reviewer-wrapper.md) — competitive framing wrapper; resolve relative to this loaded document\n","# Implementing Tasks\n\n## Overview\n\nTakes an in-memory batch of TDD-sized tasks and executes each through: implementer subagent (TDD) → PAR spec-compliance review → fix loop → PAR code-quality review with boxing-in check → fix loop → mark complete. This is a fork of `sdd` — a sibling skill in this same plugin — with the plan-file reading phase stripped and the final end-of-plan reviewer removed. Reach for `sdd` when you have a written plan file to execute; this skill is only invoked by `run-iteration`, which holds its task batch in memory.\n\n## When to Use\n\nInvoked by `run-iteration` with a list of tasks. Tasks are passed in memory, not via a file.\n\n## Validate the batch\n\nBefore dispatching anything, validate every in-memory task has a non-empty\n`Files:` field, an `Interfaces:` field, and explicit `Consumes:` and `Produces:`\nentries. `None` is the explicit value when no interface edge exists. Missing\nany field fails batch validation and returns the batch to the caller; do not\ninfer metadata from prose or execute the malformed task sequentially.\n\nFor a valid batch, use exactly two execution rungs:\n\n1. worktree-isolated parallel dispatch when files are pairwise disjoint, there\n   is no in-wave `Consumes:` → `Produces:` edge, and every worker has a\n   pairwise-unique linked Git directory;\n2. sequential dispatch of the whole group when the gate does not hold or any\n   worktree cannot be created or validated.\n\nThere is no unisolated-parallel rung. Create and validate every worktree before\ndispatching any worker so a failed setup cannot produce a partially parallel\ngroup.\n\nBefore creating a parallel group, record the controller state:\n\n```bash\nCONTROLLER_BRANCH=$(git branch --show-current)\nCONTROLLER_HEAD=$(git rev-parse HEAD)\n```\n\nEvery worker in that group must branch from exactly `CONTROLLER_HEAD`. Record\neach owned worker path and branch before dispatch. A later group cannot start\nuntil the current group's reviewed commits have been integrated into the\ncontroller.\n\n## Per-Task Cycle\n\nFor each task in the provided list:\n\n### 1. Dispatch implementer\n\nUsing the template in `implementer-subagent-prompt.md`, dispatch a single implementer subagent with:\n- The full task description and context\n- The proof obligations for each observable AC in the task's stories\n- The list of existing scenarios that may be impacted\n\nThe implementer MUST complete a pre-flight mapping (AC → proof seam → scenario) before writing code. If the implementer skips the pre-flight, re-dispatch with explicit instructions to complete it first.\n\n### 2. Handle implementer status\n\n- **DONE:** proceed to spec-compliance review (step 3). Verify the implementer's report includes pre-flight mapping and scenario updates.\n- **DONE_WITH_CONCERNS:** read the concerns. If about correctness\u002Fscope, address before review. If observations, note and proceed.\n- **NEEDS_CONTEXT:** provide the missing context and re-dispatch\n- **BLOCKED:** assess: context problem → re-dispatch with context; too hard → re-dispatch with more capable model; task too large → break into smaller pieces; plan wrong → escalate to caller\n\n### 3. PAR spec-compliance review (Stage 1)\n\nFollowing [skills\u002F_shared\u002Fparallel-adversarial-review.md](..\u002F_shared\u002Fparallel-adversarial-review.md), resolved relative to this loaded document:\n\n1. Build spec-compliance prompt using `spec-compliance-reviewer-prompt.md`\n   - Include the proof obligations and the implementer's evidence claims\n2. Resolve [skills\u002F_shared\u002Fpar-reviewer-wrapper.md](..\u002F_shared\u002Fpar-reviewer-wrapper.md) relative to this loaded document and wrap in its competitive framing\n3. Dispatch TWO spec-compliance reviewers in parallel\n4. Aggregate findings (PAR rules: union of findings, severity = take worst)\n5. If ❌ issues found:\n   - Send aggregated issues back to the implementer subagent (same subagent, via continuation message)\n   - Implementer fixes\n   - Re-dispatch fresh PAR spec-compliance pair\n   - Repeat until ✅ spec compliant with adequate evidence\n6. Only proceed to Stage 2 after Stage 1 is ✅\n\n### 4. PAR code-quality review (Stage 2)\n\nFollowing [skills\u002F_shared\u002Fparallel-adversarial-review.md](..\u002F_shared\u002Fparallel-adversarial-review.md), resolved relative to this loaded document:\n\n1. Build code-quality prompt using `code-quality-reviewer-prompt.md`\n   - Include the next 3 pending roadmap iterations for the boxing-in check\n   - Include the implementer's corpus contribution for quality review\n2. Wrap in PAR competitive framing\n3. Dispatch TWO code-quality reviewers in parallel\n4. Aggregate findings\n5. If ❌ changes needed:\n   - Send aggregated issues back to the implementer\n   - Implementer fixes\n   - Re-dispatch fresh PAR code-quality pair\n   - Repeat until ✅ approved\n\n### 5. Mark task reviewed\n\nRecord the task as reviewed. In sequential mode it is also integrated and may\nbe marked complete immediately. In parallel mode it remains pending integration\nuntil every worker in its group passes both review stages.\n\n## Integrate the Parallel Wave\n\nAfter every worker in a parallel group is reviewed:\n\n1. Return to the controller workspace and verify its branch and HEAD still\n   equal `CONTROLLER_BRANCH` and `CONTROLLER_HEAD`. If either moved, stop; a\n   concurrent change invalidated the recorded merge target.\n2. Merge each worker branch into the controller, one at a time. Stop on the\n   first conflict and preserve every worktree for investigation.\n3. Run the iteration's complete test set on the integrated controller HEAD.\n   Per-task evidence from separate branches does not prove the combined tree.\n4. When the integrated tests pass, remove each owned worker worktree, prune\n   stale registrations, and delete its merged worker branch. Never remove a\n   host-owned worktree or force removal of an unclean worktree.\n5. Mark the group's tasks complete and record the new controller HEAD. Use that\n   integrated HEAD as the base for the next group.\n\nThe controller must contain every completed task before returning to\n`run-iteration`. Returning only per-task reports while commits remain on\nworker branches is a failed iteration, not successful parallel execution.\n\nAfter all tasks are integrated, return a per-task result list to the caller, including:\n- Per-task status\n- Scenarios added or updated per task\n- Evidence commands per task\n\n## Model Selection\n\nUse the least powerful model that can handle each role:\n\n| Role | Signal → Model |\n|---|---|\n| Implementer (mechanical: 1-2 files, clear spec) | Cheap\u002Ffast model |\n| Implementer (integration: multi-file, judgment) | Standard model |\n| Spec-compliance reviewer | Standard model |\n| Code-quality reviewer | Most capable model |\n\n## Quick Reference\n\n| Per task | Subagents dispatched |\n|---|---|\n| Implementer | 1 (sequential, TDD) |\n| Spec-compliance review (PAR) | 2 in parallel |\n| Code-quality review (PAR) | 2 in parallel |\n| **Minimum per task** | **5** (before re-review loops) |\n\n## Red Flags\n\n- **Never** start code-quality review before spec compliance is ✅\n- **Never** skip the re-review after fixes (reviewer found issues = implementer fixes = review again)\n- **Do not** dispatch multiple implementers concurrently unless the batch's\n  tasks pass rung one of the worktree gate: disjoint `Files:` blocks, no\n  `Consumes:` → `Produces:` edge inside the concurrent group, and a validated,\n  pairwise-unique linked Git directory per worker (see\n  `dispatch-agents` for the full gate and `use-worktrees`\n  Step 1d). File\u002Fdependency conflicts or any worktree setup failure select rung\n  two for the whole group: sequential dispatch. Never dispatch in parallel\n  without isolation.\n- **Never** accept \"close enough\" on spec compliance\n- **Never** let implementer self-review replace the two-stage review\n\n## References\n\n- `implementer-subagent-prompt.md` — implementer dispatch template\n- `spec-compliance-reviewer-prompt.md` — Stage 1 review template\n- `code-quality-reviewer-prompt.md` — Stage 2 review template (includes boxing-in)\n- [skills\u002F_shared\u002Fparallel-adversarial-review.md](..\u002F_shared\u002Fparallel-adversarial-review.md) — PAR methodology; resolve relative to this loaded document\n- [skills\u002F_shared\u002Fpar-reviewer-wrapper.md](..\u002F_shared\u002Fpar-reviewer-wrapper.md) — competitive framing wrapper; resolve relative to this loaded document\n",{"name":5,"description":7},"main","728894f3f85fbd4e788f015ddfd72bb6155d5dd5","plugins\u002Fmoe\u002F.pi\u002Fskills\u002Fimplement-tasks",1,"https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe","https:\u002F\u002Fgithub.com\u002Fzak-keown","2026-09-05 06:29:06",{"source":8,"skillName":5,"commands":31,"prompt":41,"zipAvailable":42},[32,35,38],{"tool":33,"command":34},"npx","npx skills add zak-keown\u002Fmoe --skill implement-tasks",{"tool":36,"command":37},"bunx","bunx skills add zak-keown\u002Fmoe --skill implement-tasks",{"tool":39,"command":40},"pnpm","pnpm dlx skills add zak-keown\u002Fmoe --skill implement-tasks","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002F.pi\u002Fskills\u002Fimplement-tasks\u002FSKILL.md",true,[44,50,54],{"path":45,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":48,"sourceHtmlUrl":49},"code-quality-reviewer-prompt.md","REFERENCE","参考文档",3907,"https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002F.pi\u002Fskills\u002Fimplement-tasks\u002Fcode-quality-reviewer-prompt.md",{"path":51,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":52,"sourceHtmlUrl":53},"implementer-subagent-prompt.md",4974,"https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002F.pi\u002Fskills\u002Fimplement-tasks\u002Fimplementer-subagent-prompt.md",{"path":55,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":56,"sourceHtmlUrl":57},"spec-compliance-reviewer-prompt.md",2663,"https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002F.pi\u002Fskills\u002Fimplement-tasks\u002Fspec-compliance-reviewer-prompt.md",[59,66,73,80,87,94],{"skillNo":60,"name":61,"nameZh":6,"description":62,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":63,"categories":64,"keywords":65,"publishedAt":18},"01M7C1Z5362Y8NVRZ69HWP0ZD0","write-skill","Use when creating new skills, editing existing skills, or verifying skills work before deployment","https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002Fskills\u002Fwrite-skill\u002FSKILL.md",[],[],{"skillNo":67,"name":68,"nameZh":6,"description":69,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":70,"categories":71,"keywords":72,"publishedAt":18},"01M7C1YZ9966WK9R6ZRPCGWF3G","write-plan","Use when you have a spec or requirements for a multi-step task, before touching code","https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002Fskills\u002Fwrite-plan\u002FSKILL.md",[],[],{"skillNo":74,"name":75,"nameZh":6,"description":76,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":77,"categories":78,"keywords":79,"publishedAt":18},"01M7C1YMR4SJD7BPBBPXZCVGVH","write-clearly","Apply Strunk's timeless writing rules to ANY prose humans will read—documentation, commit messages, error messages, explanations, reports, or UI text. Makes your writing clearer, stronger, and more professional.","https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002Fskills\u002Fwrite-clearly\u002FSKILL.md",[],[],{"skillNo":81,"name":82,"nameZh":6,"description":83,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":84,"categories":85,"keywords":86,"publishedAt":18},"01M7C1YH9805YSM0N20VQAVSXF","windows-vm","Create, manage, or connect to a headless Windows 11 VM running in Docker with SSH access. Use when the user wants to spin up, stop, restart, or SSH into a Windows VM.","https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002Fskills\u002Fwindows-vm\u002FSKILL.md",[],[],{"skillNo":88,"name":89,"nameZh":6,"description":90,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":91,"categories":92,"keywords":93,"publishedAt":18},"01M7C1YD87SCNEJ2C9B8HFR2NQ","verify-completion","Use when about to claim work is complete, fixed, or passing, before committing or creating MRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always","https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002Fskills\u002Fverify-completion\u002FSKILL.md",[],[],{"skillNo":95,"name":96,"nameZh":6,"description":97,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":98,"categories":99,"keywords":100,"publishedAt":18},"01M7C1YARDEHT0TJ9BRG30ZCF4","using-moe","Use when starting any conversation - establishes how to find and use skills, requiring skill invocation before ANY response including clarifying questions","https:\u002F\u002Fgithub.com\u002Fzak-keown\u002Fmoe\u002Fblob\u002Fmain\u002Fplugins\u002Fmoe\u002Fskills\u002Fusing-moe\u002FSKILL.md",[],[]]