[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M79GYM511FZ5XSS0K536HHW1":3,"skill-same:01M79GYM511FZ5XSS0K536HHW1":49},{"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":28,"commitSha":29,"skillDirectory":30,"forksCount":14,"repositoryHtmlUrl":31,"repositoryHomepageUrl":6,"ownerHtmlUrl":32,"ownerBlogUrl":6,"ownerDescription":33,"updatedAt":34,"install":35,"resources":48},"01M79GYM511FZ5XSS0K536HHW1","long-running-task",null,"Drive a long-running task through workspace progress files (.cool\u002Ftask\u002Fprogress.md + features.json) that survive context compaction and new runs","luckystrker\u002Fcool-ai-harness","luckystrker","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F56731491?v=4","01M79GXYM0QNVKN6VMQ2QZ3GAS","Danil Kondratiuk","01M79GXYMMEYA2DMWDHE3NH982",0,"MIT","https:\u002F\u002Fgithub.com\u002Fluckystrker\u002Fcool-ai-harness\u002Fblob\u002Fmain\u002Fskills\u002Flong-running-task\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: long-running-task\ndescription: Drive a long-running task through workspace progress files (.cool\u002Ftask\u002Fprogress.md + features.json) that survive context compaction and new runs\nmetadata:\n  cool.version: \"2.0\"\n  cool.tags: \"long-running task progress checkpoint resume compaction durable\"\nallowed-tools: \"read_file write_file edit_file list_files shell git\"\n---\n\n# Long-Running Task\n\nYou are now operating in **Long-Running Task** mode. A task too big for one\ncontext window is tracked on disk so progress survives compaction, crashes\nand fresh runs.\n\n## Progress files\n\nMaintain two files in the workspace:\n\n- `.cool\u002Ftask\u002Fprogress.md` — human-readable task journal with three\n  sections:\n  - **Done** — completed work, one bullet per verified step.\n  - **Next** — the immediate next actions, ordered.\n  - **Acceptance criteria** — how the finished task is verified.\n- `.cool\u002Ftask\u002Ffeatures.json` — machine-readable checklist:\n  `{\"features\": [{\"id\": \"f1\", \"title\": \"...\", \"status\": \"pending|in_progress|done\", \"notes\": \"...\"}]}`.\n\nKeep both files consistent with each other.\n\n## Process\n\n1. **Start**: if `.cool\u002Ftask\u002Fprogress.md` exists, read it and\n   `features.json`, then resume from **Next**. Otherwise create both files:\n   break the request into a feature checklist and record acceptance\n   criteria.\n2. **Work**: do the next pending item only. Prefer `edit_file` for targeted\n   changes so progress-file updates stay small.\n3. **Update before answering**: before every reply to the user, update\n   `progress.md` (Done\u002FNext) and `features.json` statuses. The file is the\n   source of truth for the next run — a new session must be able to resume\n   from it alone.\n4. **Finish**: when every feature is `done` and the acceptance criteria are\n   met, say so explicitly and leave the final state in the files.\n\n## Principles\n\n- **File over memory**: never keep task state only in the conversation —\n  compaction may drop it. Write progress down as it happens.\n- **Small verifiable steps**: each feature should be independently\n  checkable (a test, a command, a file to inspect).\n- **State the state**: every reply briefly notes what is done and what is\n  next, matching the files.\n","# Long-Running Task\n\nYou are now operating in **Long-Running Task** mode. A task too big for one\ncontext window is tracked on disk so progress survives compaction, crashes\nand fresh runs.\n\n## Progress files\n\nMaintain two files in the workspace:\n\n- `.cool\u002Ftask\u002Fprogress.md` — human-readable task journal with three\n  sections:\n  - **Done** — completed work, one bullet per verified step.\n  - **Next** — the immediate next actions, ordered.\n  - **Acceptance criteria** — how the finished task is verified.\n- `.cool\u002Ftask\u002Ffeatures.json` — machine-readable checklist:\n  `{\"features\": [{\"id\": \"f1\", \"title\": \"...\", \"status\": \"pending|in_progress|done\", \"notes\": \"...\"}]}`.\n\nKeep both files consistent with each other.\n\n## Process\n\n1. **Start**: if `.cool\u002Ftask\u002Fprogress.md` exists, read it and\n   `features.json`, then resume from **Next**. Otherwise create both files:\n   break the request into a feature checklist and record acceptance\n   criteria.\n2. **Work**: do the next pending item only. Prefer `edit_file` for targeted\n   changes so progress-file updates stay small.\n3. **Update before answering**: before every reply to the user, update\n   `progress.md` (Done\u002FNext) and `features.json` statuses. The file is the\n   source of truth for the next run — a new session must be able to resume\n   from it alone.\n4. **Finish**: when every feature is `done` and the acceptance criteria are\n   met, say so explicitly and leave the final state in the files.\n\n## Principles\n\n- **File over memory**: never keep task state only in the conversation —\n  compaction may drop it. Write progress down as it happens.\n- **Small verifiable steps**: each feature should be independently\n  checkable (a test, a command, a file to inspect).\n- **State the state**: every reply briefly notes what is done and what is\n  next, matching the files.\n",{"name":5,"metadata":24,"description":7,"allowed-tools":27},{"cool.tags":25,"cool.version":26},"long-running task progress checkpoint resume compaction durable","2.0","read_file write_file edit_file list_files shell git","main","d4630c1a106a3120a59060e0af068ee0eb7244e8","skills\u002Flong-running-task","https:\u002F\u002Fgithub.com\u002Fluckystrker\u002Fcool-ai-harness","https:\u002F\u002Fgithub.com\u002Fluckystrker","- 24 y\u002Fo \r\n- Software Engineer\r\n- AI Evangelist -","2026-11-03 09:26:29",{"source":8,"skillName":5,"commands":36,"prompt":46,"zipAvailable":47},[37,40,43],{"tool":38,"command":39},"npx","npx skills add luckystrker\u002Fcool-ai-harness --skill long-running-task",{"tool":41,"command":42},"bunx","bunx skills add luckystrker\u002Fcool-ai-harness --skill long-running-task",{"tool":44,"command":45},"pnpm","pnpm dlx skills add luckystrker\u002Fcool-ai-harness --skill long-running-task","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fluckystrker\u002Fcool-ai-harness\u002Fblob\u002Fmain\u002Fskills\u002Flong-running-task\u002FSKILL.md",true,[],[50,57,64,71,78,85],{"skillNo":51,"name":52,"nameZh":6,"description":53,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":54,"categories":55,"keywords":56,"publishedAt":19},"01M79GYWJSENZEBQGXZJ1F6SCT","translate","Translate text between languages while preserving meaning, tone, and formatting","https:\u002F\u002Fgithub.com\u002Fluckystrker\u002Fcool-ai-harness\u002Fblob\u002Fmain\u002Fskills\u002Ftranslate\u002FSKILL.md",[],[],{"skillNo":58,"name":59,"nameZh":6,"description":60,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":61,"categories":62,"keywords":63,"publishedAt":19},"01M79GYSA18J9N5551942K3HKV","summarize-document","Summarize documents, articles, or long text into concise structured summaries","https:\u002F\u002Fgithub.com\u002Fluckystrker\u002Fcool-ai-harness\u002Fblob\u002Fmain\u002Fskills\u002Fsummarize-document\u002FSKILL.md",[],[],{"skillNo":65,"name":66,"nameZh":6,"description":67,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":68,"categories":69,"keywords":70,"publishedAt":19},"01M79GYPE2SEC49VYMKWFE6SK1","skill-creation","Create new skills based on user requests - generates properly formatted SKILL.md files with metadata and instructions","https:\u002F\u002Fgithub.com\u002Fluckystrker\u002Fcool-ai-harness\u002Fblob\u002Fmain\u002Fskills\u002Fskill-creation\u002FSKILL.md",[],[],{"skillNo":72,"name":73,"nameZh":6,"description":74,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":75,"categories":76,"keywords":77,"publishedAt":19},"01M79GYH2KCPZJY1KV6A1JSN2V","deep-research","Conduct thorough multi-source research on a topic, synthesizing findings into a comprehensive report","https:\u002F\u002Fgithub.com\u002Fluckystrker\u002Fcool-ai-harness\u002Fblob\u002Fmain\u002Fskills\u002Fdeep-research\u002FSKILL.md",[],[],{"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":19},"01M79GYESNH4AP6NEDCVWRSR9J","code-task","Execute a structured coding task with best practices - read, plan, implement, test, commit","https:\u002F\u002Fgithub.com\u002Fluckystrker\u002Fcool-ai-harness\u002Fblob\u002Fmain\u002Fskills\u002Fcode-task\u002FSKILL.md",[],[],{"skillNo":86,"name":87,"nameZh":6,"description":88,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":89,"categories":90,"keywords":91,"publishedAt":19},"01M79GYCXGBPYD5JV4E4HPNP60","brainstorm","Generate creative ideas, explore possibilities, and structure thinking around a problem or opportunity","https:\u002F\u002Fgithub.com\u002Fluckystrker\u002Fcool-ai-harness\u002Fblob\u002Fmain\u002Fskills\u002Fbrainstorm\u002FSKILL.md",[],[]]