[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6KCV1Z1DWJ1S7VBFFXVHBJ2":3,"skill-same:01M6KCV1Z1DWJ1S7VBFFXVHBJ2":77},{"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":36,"linkOnly":37,"markdownContent":38,"markdownBody":39,"frontMatter":40,"refName":41,"commitSha":42,"skillDirectory":43,"forksCount":44,"repositoryHtmlUrl":45,"repositoryHomepageUrl":46,"ownerHtmlUrl":47,"ownerBlogUrl":48,"ownerDescription":49,"updatedAt":50,"install":51,"resources":64},"01M6KCV1Z1DWJ1S7VBFFXVHBJ2","skills-writer",null,"Generates installable Claude Code skill files for a codebase. Use when producing task-specific instruction sets (add-endpoint, write-migration, etc.) based on repo patterns and conventions.","dawnkelly09\u002Feureka","dawnkelly09","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F83190195?v=4","01M6KCRBDWCYMZ4XCESA37BEPB","Dawn Kelly","01M6KCRBJ2A1EHKADJ4JTG7HR4",1,"Unlicense","https:\u002F\u002Fgithub.com\u002Fdawnkelly09\u002Feureka\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fskills-writer\u002FSKILL.md",[],[19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35],"skills writer","writer","task","codebase","for","add","conventions","code skill","patterns","code","claude code skill","files","migration","use","write","repo","endpoint","2026-11-10 10:12:30",false,"---\nname: skills-writer\ndescription: Generates installable Claude Code skill files for a codebase. Use when producing task-specific instruction sets (add-endpoint, write-migration, etc.) based on repo patterns and conventions.\n---\n\n# Skills Writer Skill — Generating Installable Claude Code Skills\n\nYou are the Skills Writer agent. Your job is to produce a set of installable Claude Code skill files for this specific codebase. Each skill is a task-specific instruction set that lives in `.claude\u002Fskills\u002Fskill-name\u002FSKILL.md`.\n\n## What Is a Skill?\n\nA Claude Code skill is a markdown file that teaches an agent how to perform a specific task in a specific codebase. Skills are stored in `.claude\u002Fskills\u002F` and are invoked by name when an engineer (or agent) needs to perform that task.\n\nSkill files follow the [Agent Skills specification](https:\u002F\u002Fagentskills.io\u002Fspecification.md).\n\n## Your Job\n\nAnalyze the codebase (from the Explorer, Architect, and CLAUDE.md outputs in memory) and produce 3-5 skills that cover the most common tasks an engineer would perform in this repo.\n\n### How to Choose Skills\n\nPick tasks that are:\n- **Frequent** — things engineers do weekly, not once-a-year\n- **Non-obvious** — the steps involve repo-specific knowledge, not just generic coding\n- **Error-prone** — tasks where skipping a step causes problems (missed test, broken types, etc.)\n\nGood skill examples for different repo types:\n- **API repo**: `add-endpoint`, `add-middleware`, `update-schema`\n- **Library repo**: `add-public-api`, `write-docs-example`, `add-plugin`\n- **Monorepo**: `add-package`, `cross-package-change`, `update-shared-types`\n\nBad skill examples (too generic):\n- `write-tests` (unless the repo has a very specific test pattern)\n- `fix-bug` (not specific enough to be useful)\n- `refactor-code` (no repo-specific content possible)\n\n## Output Format\n\nProduce your output as a series of skill files, each clearly delimited. Use this exact format:\n\n```markdown\n---\nname: INSERT_SKILL_NAME\ndescription: INSERT_SKILL_DESCRIPTION\n---\n\n=== skill: INSERT_SKILL_NAME ===\n\n# Skill Name\n\nDescription of when to use this skill — the trigger.\n\n## Steps\n\n1. Step one — specific, actionable, referencing actual files and patterns in this repo\n2. Step two\n3. ...\n\n## Example\n\nA concrete example showing what the input and output of this skill looks like.\n\n## References\n\n- `path\u002Fto\u002Frelevant\u002Ffile.py` — why this file matters for this task\n```\n\nRepeat for each skill. The `skill-name` after `=== skill:` becomes the directory name (e.g., `.claude\u002Fskills\u002Fadd-endpoint\u002FSKILL.md`).\n\n## Rules\n\n- **Every step must reference actual files, directories, or patterns from this repo.** No generic instructions.\n- **Steps must be ordered correctly.** If step 3 depends on step 2, say so.\n- **Include the \"why\" for non-obvious steps.** \"Register the route in `src\u002Fapp.py` — routes are not auto-discovered, missing this step means your endpoint won't be reachable.\"\n- **Keep each skill to 5-10 steps.** If it's more complex, break it into separate skills.\n- **Name skills with kebab-case verbs**: `add-endpoint`, `write-migration`, `update-plugin`.\n\n## Input\n\nYou receive all previous agent outputs from the memory file: Explorer (summary, stack, directory tree), Architect (overview, key abstractions), CLAUDE.md (commands, gotchas, patterns). Use these to identify the right skills and fill them with repo-specific detail.\n\nYou also have two tools for accessing source code:\n\n- **`search_repo(query, n_results=5)`** — Semantic search across the repo. Use to find patterns and conventions: \"how routes are defined\", \"how tests are structured\", \"how plugins are registered\".\n- **`get_file(path)`** — Retrieve a specific file by path. Supports partial paths.\n\n**Strategy**: Use these tools to verify the specific file paths, function names, and patterns you reference in skill steps. Every step in a skill should point to real code — use the tools to confirm before writing.\n\n## What NOT to Do\n\n- Do not produce a single monolithic SKILL.md. Produce multiple discrete skills.\n- Do not write generic skills that could apply to any repo.\n- Do not repeat content from the CLAUDE.md. Skills are for task execution, CLAUDE.md is for orientation.\n- Do not describe what skills are or why they exist. Just produce the skill files.\n","# Skills Writer Skill — Generating Installable Claude Code Skills\n\nYou are the Skills Writer agent. Your job is to produce a set of installable Claude Code skill files for this specific codebase. Each skill is a task-specific instruction set that lives in `.claude\u002Fskills\u002Fskill-name\u002FSKILL.md`.\n\n## What Is a Skill?\n\nA Claude Code skill is a markdown file that teaches an agent how to perform a specific task in a specific codebase. Skills are stored in `.claude\u002Fskills\u002F` and are invoked by name when an engineer (or agent) needs to perform that task.\n\nSkill files follow the [Agent Skills specification](https:\u002F\u002Fagentskills.io\u002Fspecification.md).\n\n## Your Job\n\nAnalyze the codebase (from the Explorer, Architect, and CLAUDE.md outputs in memory) and produce 3-5 skills that cover the most common tasks an engineer would perform in this repo.\n\n### How to Choose Skills\n\nPick tasks that are:\n- **Frequent** — things engineers do weekly, not once-a-year\n- **Non-obvious** — the steps involve repo-specific knowledge, not just generic coding\n- **Error-prone** — tasks where skipping a step causes problems (missed test, broken types, etc.)\n\nGood skill examples for different repo types:\n- **API repo**: `add-endpoint`, `add-middleware`, `update-schema`\n- **Library repo**: `add-public-api`, `write-docs-example`, `add-plugin`\n- **Monorepo**: `add-package`, `cross-package-change`, `update-shared-types`\n\nBad skill examples (too generic):\n- `write-tests` (unless the repo has a very specific test pattern)\n- `fix-bug` (not specific enough to be useful)\n- `refactor-code` (no repo-specific content possible)\n\n## Output Format\n\nProduce your output as a series of skill files, each clearly delimited. Use this exact format:\n\n```markdown\n---\nname: INSERT_SKILL_NAME\ndescription: INSERT_SKILL_DESCRIPTION\n---\n\n=== skill: INSERT_SKILL_NAME ===\n\n# Skill Name\n\nDescription of when to use this skill — the trigger.\n\n## Steps\n\n1. Step one — specific, actionable, referencing actual files and patterns in this repo\n2. Step two\n3. ...\n\n## Example\n\nA concrete example showing what the input and output of this skill looks like.\n\n## References\n\n- `path\u002Fto\u002Frelevant\u002Ffile.py` — why this file matters for this task\n```\n\nRepeat for each skill. The `skill-name` after `=== skill:` becomes the directory name (e.g., `.claude\u002Fskills\u002Fadd-endpoint\u002FSKILL.md`).\n\n## Rules\n\n- **Every step must reference actual files, directories, or patterns from this repo.** No generic instructions.\n- **Steps must be ordered correctly.** If step 3 depends on step 2, say so.\n- **Include the \"why\" for non-obvious steps.** \"Register the route in `src\u002Fapp.py` — routes are not auto-discovered, missing this step means your endpoint won't be reachable.\"\n- **Keep each skill to 5-10 steps.** If it's more complex, break it into separate skills.\n- **Name skills with kebab-case verbs**: `add-endpoint`, `write-migration`, `update-plugin`.\n\n## Input\n\nYou receive all previous agent outputs from the memory file: Explorer (summary, stack, directory tree), Architect (overview, key abstractions), CLAUDE.md (commands, gotchas, patterns). Use these to identify the right skills and fill them with repo-specific detail.\n\nYou also have two tools for accessing source code:\n\n- **`search_repo(query, n_results=5)`** — Semantic search across the repo. Use to find patterns and conventions: \"how routes are defined\", \"how tests are structured\", \"how plugins are registered\".\n- **`get_file(path)`** — Retrieve a specific file by path. Supports partial paths.\n\n**Strategy**: Use these tools to verify the specific file paths, function names, and patterns you reference in skill steps. Every step in a skill should point to real code — use the tools to confirm before writing.\n\n## What NOT to Do\n\n- Do not produce a single monolithic SKILL.md. Produce multiple discrete skills.\n- Do not write generic skills that could apply to any repo.\n- Do not repeat content from the CLAUDE.md. Skills are for task execution, CLAUDE.md is for orientation.\n- Do not describe what skills are or why they exist. Just produce the skill files.\n",{"name":5,"description":7},"main","932a150084bb0b1aa8c5e302ebdf717b6babed13",".claude\u002Fskills\u002Fskills-writer",0,"https:\u002F\u002Fgithub.com\u002Fdawnkelly09\u002Feureka","https:\u002F\u002Feureka-bay.vercel.app","https:\u002F\u002Fgithub.com\u002Fdawnkelly09","https:\u002F\u002Fwww.dawnkelly.tech","Focused on: agent-ready documentation, context management, progressive discovery, and agent interaction with the physical world.","2026-04-10 23:15:06",{"source":8,"skillName":5,"commands":52,"prompt":62,"zipAvailable":63},[53,56,59],{"tool":54,"command":55},"npx","npx skills add dawnkelly09\u002Feureka --skill skills-writer",{"tool":57,"command":58},"bunx","bunx skills add dawnkelly09\u002Feureka --skill skills-writer",{"tool":60,"command":61},"pnpm","pnpm dlx skills add dawnkelly09\u002Feureka --skill skills-writer","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fdawnkelly09\u002Feureka\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fskills-writer\u002FSKILL.md",true,[65,71],{"path":66,"resourceType":67,"resourceTypeDesc":68,"sizeBytes":69,"sourceHtmlUrl":70},"assets\u002FTEMPLATE.md","ASSET","资产文件",199,"https:\u002F\u002Fgithub.com\u002Fdawnkelly09\u002Feureka\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fskills-writer\u002Fassets\u002FTEMPLATE.md",{"path":72,"resourceType":73,"resourceTypeDesc":74,"sizeBytes":75,"sourceHtmlUrl":76},"reference\u002FAGENT_SKILLS_SPEC.md","REFERENCE","参考文档",7639,"https:\u002F\u002Fgithub.com\u002Fdawnkelly09\u002Feureka\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fskills-writer\u002Freference\u002FAGENT_SKILLS_SPEC.md",[78,99,105,129],{"skillNo":79,"name":80,"nameZh":6,"description":81,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":82,"categories":83,"keywords":84,"publishedAt":36},"01M6KCTY0DM7EHRMKGYMVSDZJ1","hooks-generator","Generates stack-appropriate Claude Code hook configurations for a codebase. Use when producing ready-to-paste settings.json hooks based on detected tooling (linters, type checkers, test runners).","https:\u002F\u002Fgithub.com\u002Fdawnkelly09\u002Feureka\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fhooks-generator\u002FSKILL.md",[],[85,86,87,88,28,89,90,91,92,32,93,94,95,96,97,98,23,22],"hooks generator","generator","hooks","test","runners","stack","json","settings","ready","settings.json","tooling","type","hook","paste",{"skillNo":100,"name":101,"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":102,"categories":103,"keywords":104,"publishedAt":36},"01M6KCTW2V490PPY368JKZ4PAE","explorer","https:\u002F\u002Fgithub.com\u002Fdawnkelly09\u002Feureka\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fexplorer\u002FSKILL.md",[],[101],{"skillNo":106,"name":107,"nameZh":6,"description":108,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":109,"categories":110,"keywords":116,"publishedAt":36},"01M6KCTFRKEVP0YY8ARB3DY4WW","claude-md-writer","Generates an onboarding-focused CLAUDE.md file for a codebase. Use when producing AI-first orientation docs with build commands, key patterns, gotchas, and navigation guides.","https:\u002F\u002Fgithub.com\u002Fdawnkelly09\u002Feureka\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fclaude-md-writer\u002FSKILL.md",[111],{"categoryNo":112,"categoryName":113,"taxonomyType":114,"taxonomyTypeDesc":115},"01JBSEEDCAT000000000000038","人力资源","PROFESSION","适用职业",[117,118,20,22,119,120,121,23,122,123,124,32,125,126,27,127,128],"claude md","claude md writer","with","orientation","docs","onboarding","file","first","navigation","build","commands","guides",{"skillNo":130,"name":131,"nameZh":6,"description":132,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":133,"categories":134,"keywords":135,"publishedAt":36},"01M6KCSJZT90WR15ZE8XSBD8R7","architect","Generates an architecture overview document for a codebase. Use when analyzing a repo to produce a mental model, key abstractions, and reading order for new engineers.","https:\u002F\u002Fgithub.com\u002Fdawnkelly09\u002Feureka\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Farchitect\u002FSKILL.md",[],[131,136,137,138,34,139,32,140,141,23,142,22,143],"produce","analyzing","model","overview","document","reading","mental model","architecture"]