[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6CS49SJ1BHSCN8Q8TEWM7K9":3,"skill-same:01M6CS49SJ1BHSCN8Q8TEWM7K9":97},{"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":26,"publishedAt":45,"linkOnly":46,"markdownContent":47,"markdownBody":48,"frontMatter":49,"refName":50,"commitSha":51,"skillDirectory":52,"forksCount":53,"repositoryHtmlUrl":54,"repositoryHomepageUrl":6,"ownerHtmlUrl":55,"ownerBlogUrl":56,"ownerDescription":57,"updatedAt":58,"install":59,"resources":72},"01M6CS49SJ1BHSCN8Q8TEWM7K9","deep-research",null,"Read when the OPERATOR asks a question that will require web research across multiple sources — recommendations, comparisons, evaluations, multi-factor decisions, \"what should I buy\u002Fuse\", or any question where you'd need to read several web pages. This skill decides whether to spawn a background research agent (to protect your context from heavy fetching) or handle it yourself. Do NOT read for simple factual lookups or questions fully answered by your knowledge base.","mrtolkien\u002FGHOST","mrtolkien","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F8489650?v=4","01M6CS32Q934XBQQE8BDV5Z6W2","Tolki","01M6CS32VBCD47F5E27WF9W6FC",5,"MIT","https:\u002F\u002Fgithub.com\u002Fmrtolkien\u002FGHOST\u002Fblob\u002Fmain\u002Fassets\u002Fskills\u002Fdeep-research\u002Fskill.md",[18,23],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000021","资料检索与研究","SCENARIO","使用场景",{"categoryNo":24,"categoryName":25,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000023","Skill与智能体开发",[27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44],"deep research","research","question","base","do","protect","simple","use","spawn","context","read","decisions","web research","knowledge","research agent","multi-factor","operator","web","2026-11-10 10:12:30",false,"---\nname: deep-research\ndescription:\n  Read when the OPERATOR asks a question that will require web research across multiple\n  sources — recommendations, comparisons, evaluations, multi-factor decisions, \"what\n  should I buy\u002Fuse\", or any question where you'd need to read several web pages. This\n  skill decides whether to spawn a background research agent (to protect your context\n  from heavy fetching) or handle it yourself. Do NOT read for simple factual lookups or\n  questions fully answered by your knowledge base.\n---\n\n# Deep Research Skill\n\nYou're reading this because the OPERATOR's question needs multi-source research.\n\n## Why the Agent Exists\n\nEach `web_fetch` dumps thousands of tokens into your context. Doing several fetches\ninline to answer one question pollutes your main conversation — past messages get\ncompressed, future turns get worse. The deep-research agent runs in an **isolated\ncontext** that is discarded after it delivers a summary. It protects your conversation\nwhile doing the heavy reading.\n\n## Decision Process\n\n### Step 1: Check knowledge\n\nCall `knowledge_search` first. If you find existing notes or references that adequately\nanswer the question, use them and respond directly. No agent needed.\n\n### Step 1.5: Consider reference import first\n\nIf the research centers on a **specific tool, framework, or library** (e.g. Ergogen,\nDioxus, KiCad), check whether its official docs can be imported via the\n`reference-import` skill. Imported docs persist in your knowledge base and are\nsearchable across sessions — far more valuable than ephemeral agent context. Read the\nreference-import skill, start the import in the background, and _then_ optionally spawn\nthe deep-research agent for the broader question (e.g. community guides, best practices)\nwhile the import runs.\n\n### Step 2: Spawn the agent\n\nIf knowledge didn't have a good answer (and reference import doesn't apply or is already\nrunning), spawn the deep-research agent. You matched this skill's description because\nthe question needs multi-source research — that research belongs in the agent's isolated\ncontext, not inline.\n\n**Your next tool call after the knowledge check must be `agent`.** Do not call\n`web_search` or `web_fetch` — every page you fetch inline is context you can never\nreclaim. Let the agent do the heavy reading.\n\n## Spawning the Agent\n\nUse `agent(action: 'start', name: 'deep-research', prompt: '...')`. Include:\n\n- **Specific question** — what exactly needs to be answered\n- **Context** — constraints, preferences, use case the OPERATOR mentioned\n- **Scope** — what sub-questions to investigate\n- **Recency** — remind the agent to look for recent developments\n- **Known sources** — if your knowledge base has source quality notes for this domain,\n  pass them to the agent so it can skip the source-discovery phase\n- **Don't name specific options** — describe the category, constraints, and use case,\n  but do NOT pre-populate specific names, brands, or recommendations from your training\n  data. Your knowledge is outdated — let the agent discover the current landscape\n  through research. Bad: \"looking at X, Y, and Z\". Good: \"looking for a [category] that\n  meets [constraints]\".\n\nAfter spawning, tell the OPERATOR you've started a background research task and you'll\nshare findings when it completes.\n\n## Follow-Up Questions\n\nWhen the OPERATOR provides follow-up criteria or refinements after the agent delivered\nfindings, **continue** the existing agent session:\n\n```\nagent(action: 'continue', agent_id: '\u003Cid>',\n  prompt: '\u003Cnew constraints and follow-up question>')\n```\n\nThe agent already has full context from its research. Continuing lets it do targeted\nfollow-up searches instead of starting from scratch. Never spawn a new agent for\nfollow-ups on the same topic.\n","# Deep Research Skill\n\nYou're reading this because the OPERATOR's question needs multi-source research.\n\n## Why the Agent Exists\n\nEach `web_fetch` dumps thousands of tokens into your context. Doing several fetches\ninline to answer one question pollutes your main conversation — past messages get\ncompressed, future turns get worse. The deep-research agent runs in an **isolated\ncontext** that is discarded after it delivers a summary. It protects your conversation\nwhile doing the heavy reading.\n\n## Decision Process\n\n### Step 1: Check knowledge\n\nCall `knowledge_search` first. If you find existing notes or references that adequately\nanswer the question, use them and respond directly. No agent needed.\n\n### Step 1.5: Consider reference import first\n\nIf the research centers on a **specific tool, framework, or library** (e.g. Ergogen,\nDioxus, KiCad), check whether its official docs can be imported via the\n`reference-import` skill. Imported docs persist in your knowledge base and are\nsearchable across sessions — far more valuable than ephemeral agent context. Read the\nreference-import skill, start the import in the background, and _then_ optionally spawn\nthe deep-research agent for the broader question (e.g. community guides, best practices)\nwhile the import runs.\n\n### Step 2: Spawn the agent\n\nIf knowledge didn't have a good answer (and reference import doesn't apply or is already\nrunning), spawn the deep-research agent. You matched this skill's description because\nthe question needs multi-source research — that research belongs in the agent's isolated\ncontext, not inline.\n\n**Your next tool call after the knowledge check must be `agent`.** Do not call\n`web_search` or `web_fetch` — every page you fetch inline is context you can never\nreclaim. Let the agent do the heavy reading.\n\n## Spawning the Agent\n\nUse `agent(action: 'start', name: 'deep-research', prompt: '...')`. Include:\n\n- **Specific question** — what exactly needs to be answered\n- **Context** — constraints, preferences, use case the OPERATOR mentioned\n- **Scope** — what sub-questions to investigate\n- **Recency** — remind the agent to look for recent developments\n- **Known sources** — if your knowledge base has source quality notes for this domain,\n  pass them to the agent so it can skip the source-discovery phase\n- **Don't name specific options** — describe the category, constraints, and use case,\n  but do NOT pre-populate specific names, brands, or recommendations from your training\n  data. Your knowledge is outdated — let the agent discover the current landscape\n  through research. Bad: \"looking at X, Y, and Z\". Good: \"looking for a [category] that\n  meets [constraints]\".\n\nAfter spawning, tell the OPERATOR you've started a background research task and you'll\nshare findings when it completes.\n\n## Follow-Up Questions\n\nWhen the OPERATOR provides follow-up criteria or refinements after the agent delivered\nfindings, **continue** the existing agent session:\n\n```\nagent(action: 'continue', agent_id: '\u003Cid>',\n  prompt: '\u003Cnew constraints and follow-up question>')\n```\n\nThe agent already has full context from its research. Continuing lets it do targeted\nfollow-up searches instead of starting from scratch. Never spawn a new agent for\nfollow-ups on the same topic.\n",{"name":5,"description":7},"main","083d96429a2d05526f8a294b83c74f0336670def","assets\u002Fskills\u002Fdeep-research",0,"https:\u002F\u002Fgithub.com\u002Fmrtolkien\u002FGHOST","https:\u002F\u002Fgithub.com\u002Fmrtolkien","https:\u002F\u002Fblog.tolki.dev","Data Engineer \u002F Esports \u002F Keyboards","2026-04-24 15:41:08",{"source":8,"skillName":5,"commands":60,"prompt":70,"zipAvailable":71},[61,64,67],{"tool":62,"command":63},"npx","npx skills add mrtolkien\u002FGHOST --skill deep-research",{"tool":65,"command":66},"bunx","bunx skills add mrtolkien\u002FGHOST --skill deep-research",{"tool":68,"command":69},"pnpm","pnpm dlx skills add mrtolkien\u002FGHOST --skill deep-research","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fmrtolkien\u002FGHOST\u002Fblob\u002Fmain\u002Fassets\u002Fskills\u002Fdeep-research\u002Fskill.md",true,[73,79,85,89,93],{"path":74,"resourceType":75,"resourceTypeDesc":76,"sizeBytes":77,"sourceHtmlUrl":78},"deep-research-reflection\u002Fagent.lua","OTHER","其他",2346,"https:\u002F\u002Fgithub.com\u002Fmrtolkien\u002FGHOST\u002Fblob\u002Fmain\u002Fassets\u002Fskills\u002Fdeep-research\u002Fdeep-research-reflection\u002Fagent.lua",{"path":80,"resourceType":81,"resourceTypeDesc":82,"sizeBytes":83,"sourceHtmlUrl":84},"deep-research-reflection\u002Fprompt.md","REFERENCE","参考文档",3601,"https:\u002F\u002Fgithub.com\u002Fmrtolkien\u002FGHOST\u002Fblob\u002Fmain\u002Fassets\u002Fskills\u002Fdeep-research\u002Fdeep-research-reflection\u002Fprompt.md",{"path":86,"resourceType":81,"resourceTypeDesc":82,"sizeBytes":87,"sourceHtmlUrl":88},"deep-research-reflection\u002Fuser-message.md",309,"https:\u002F\u002Fgithub.com\u002Fmrtolkien\u002FGHOST\u002Fblob\u002Fmain\u002Fassets\u002Fskills\u002Fdeep-research\u002Fdeep-research-reflection\u002Fuser-message.md",{"path":90,"resourceType":75,"resourceTypeDesc":76,"sizeBytes":91,"sourceHtmlUrl":92},"deep-research\u002Fagent.lua",7153,"https:\u002F\u002Fgithub.com\u002Fmrtolkien\u002FGHOST\u002Fblob\u002Fmain\u002Fassets\u002Fskills\u002Fdeep-research\u002Fdeep-research\u002Fagent.lua",{"path":94,"resourceType":81,"resourceTypeDesc":82,"sizeBytes":95,"sourceHtmlUrl":96},"deep-research\u002Fprompt.md",5280,"https:\u002F\u002Fgithub.com\u002Fmrtolkien\u002FGHOST\u002Fblob\u002Fmain\u002Fassets\u002Fskills\u002Fdeep-research\u002Fdeep-research\u002Fprompt.md",[98,124,148,165,174,185],{"skillNo":99,"name":100,"nameZh":6,"description":101,"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":45},"01M6CS613RWD9CFJ3KFX47V1DS","system-management","Use when reading or changing GHOST configuration (providers, models, timing, compaction, services, any config.toml key), adding or switching LLM providers, answering questions about GHOST's own features or capabilities, a CLI tool is missing or \"not found\", services need starting\u002Fstopping\u002Fdebugging, updating GHOST to a newer version, managing the Nix shell environment, or troubleshooting GHOST setup.","https:\u002F\u002Fgithub.com\u002Fmrtolkien\u002FGHOST\u002Fblob\u002Fmain\u002Fassets\u002Fskills\u002Fsystem-management\u002Fskill.md",[],[105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,34,121,122,123],"system management","management","system","shell","cli","nix","config","setup","debugging","capabilities","services","Providers","ghost","tool","troubleshooting","environment","version","models","configuration",{"skillNo":125,"name":126,"nameZh":6,"description":127,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":128,"categories":129,"keywords":141,"publishedAt":45},"01M6CS5XTK2884R9HE15YR8YZM","writing-skills","Use when creating new skills, editing existing skills, or verifying skills work before deployment","https:\u002F\u002Fgithub.com\u002Fmrtolkien\u002FGHOST\u002Fblob\u002Fmain\u002Fassets\u002Fskills\u002Fsuperpowers\u002Fwriting-skills\u002Fskill.md",[130,133,136],{"categoryNo":131,"categoryName":132,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000008","测试与质量",{"categoryNo":134,"categoryName":135,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000018","写作与文案",{"categoryNo":137,"categoryName":138,"taxonomyType":139,"taxonomyTypeDesc":140},"01JBSEEDCAT000000000000036","内容创作者","PROFESSION","适用职业",[142,143,144,145,146,34,147],"writing skills","writing","coding","work","editing","deployment",{"skillNo":149,"name":150,"nameZh":6,"description":151,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":152,"categories":153,"keywords":156,"publishedAt":45},"01M6CS5V737R1DVB3TZEK0C6WT","writing-plans","Use when you have a spec or requirements for a multi-step task, before touching code. Write comprehensive implementation plans assuming the engineer has zero context for our codebase and questionable taste.","https:\u002F\u002Fgithub.com\u002Fmrtolkien\u002FGHOST\u002Fblob\u002Fmain\u002Fassets\u002Fskills\u002Fsuperpowers\u002Fwriting-plans\u002Fskill.md",[154,155],{"categoryNo":134,"categoryName":135,"taxonomyType":21,"taxonomyTypeDesc":22},{"categoryNo":137,"categoryName":138,"taxonomyType":139,"taxonomyTypeDesc":140},[157,143,158,159,160,144,161,36,34,162,163,164],"writing plans","spec","code","ZeRO","write","engineer","implementation","task",{"skillNo":166,"name":167,"nameZh":6,"description":168,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":169,"categories":170,"keywords":171,"publishedAt":45},"01M6CS5S5S0YY56686G5CQ0WMP","verification","Use when about to claim work is complete, fixed, or passing, before committing or creating PRs. Requires running verification commands and confirming output before making any success claims. Evidence before assertions, always.","https:\u002F\u002Fgithub.com\u002Fmrtolkien\u002FGHOST\u002Fblob\u002Fmain\u002Fassets\u002Fskills\u002Fsuperpowers\u002Fverification\u002Fskill.md",[],[167,144,145,34,172,173],"evidence","claim",{"skillNo":175,"name":176,"nameZh":6,"description":177,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":178,"categories":179,"keywords":181,"publishedAt":45},"01M6CS5PD2ZSFWYRDZSACNTAJY","tdd","Use when implementing any feature or bugfix, before writing implementation code. Write the test first. Watch it fail. Write minimal code to pass.","https:\u002F\u002Fgithub.com\u002Fmrtolkien\u002FGHOST\u002Fblob\u002Fmain\u002Fassets\u002Fskills\u002Fsuperpowers\u002Ftdd\u002Fskill.md",[180],{"categoryNo":131,"categoryName":132,"taxonomyType":21,"taxonomyTypeDesc":22},[176,143,159,182,144,161,34,183,163,184],"first","minimal","test",{"skillNo":186,"name":187,"nameZh":6,"description":188,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":189,"categories":190,"keywords":192,"publishedAt":45},"01M6CS5K5WE2SHF1YV8D9C4C77","systematic-debugging","Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes","https:\u002F\u002Fgithub.com\u002Fmrtolkien\u002FGHOST\u002Fblob\u002Fmain\u002Fassets\u002Fskills\u002Fsuperpowers\u002Fsystematic-debugging\u002Fskill.md",[191],{"categoryNo":131,"categoryName":132,"taxonomyType":21,"taxonomyTypeDesc":22},[193,113,194,34,144,184,195],"systematic debugging","behavior","bug"]