[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7CYX7V3CEGDMK16ER3PSW1J":3,"skill-same:01M7CYX7V3CEGDMK16ER3PSW1J":43},{"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":13,"repositoryHtmlUrl":26,"repositoryHomepageUrl":6,"ownerHtmlUrl":27,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":28,"install":29,"resources":42},"01M7CYX7V3CEGDMK16ER3PSW1J","research",null,"Research a topic exhaustively before planning or building — discuss the goal, draft a comprehensive question set, dispatch parallel research agents against it, and synthesize a findings report. Stops there; does NOT chain into \u002Fgrill-me or plan mode. Use when the user wants to research, investigate, or survey a topic\u002Ftechnology\u002Fapproach before deciding what to do about it.","jpol34\u002Fpolyak-dev-kit","jpol34","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F14319666?v=4","01M7CYWYBNQWVRNN6SNRX8V8B8","01M7CYWYCQZX9312AEP87MYQ42",0,"MIT","https:\u002F\u002Fgithub.com\u002Fjpol34\u002Fpolyak-dev-kit\u002Fblob\u002Fmaster\u002Fskills\u002Fresearch\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: research\ndescription: Research a topic exhaustively before planning or building — discuss the goal, draft a comprehensive question set, dispatch parallel research agents against it, and synthesize a findings report. Stops there; does NOT chain into \u002Fgrill-me or plan mode. Use when the user wants to research, investigate, or survey a topic\u002Ftechnology\u002Fapproach before deciding what to do about it.\ndisable-model-invocation: false\n---\n\n# Research\n\nTurn a loose topic into a synthesized, sourced findings report. This is the step *before*\n`\u002Fgrill-me` and plan mode, not a replacement for them — it ends by handing the user a report,\nnever by auto-continuing into an interview or a plan.\n\n## Process\n\n### 1. Discuss\n\nGet the topic and the reason it matters in a sentence or two: what decision or question this\nresearch is meant to inform. If the user already stated it, don't re-ask — just confirm you have\nenough to scope the question set. If it's genuinely vague (no worked-out angle at all), ask one\nclarifying question rather than guessing at scope; don't turn this into a `\u002Fgrill-me` interview.\n\n### 2. Draft the question set\n\nWrite an exhaustive set of sub-questions covering the topic, grouped by theme (e.g. prior art \u002F\nexisting solutions, technical tradeoffs, cost\u002Fpricing, risks and failure modes, ecosystem\u002Fvendor\nlandscape, what would change the answer). Aim for coverage, not padding — skip themes that don't\napply to this topic rather than forcing a fixed template.\n\nShow the numbered, grouped list to the user before dispatching anything. This is a cheap\ncheckpoint: dispatching a dozen parallel agents on the wrong question set is the expensive\nmistake it prevents. Let the user add, cut, or re-scope questions. Skip the pause only if the user\nexplicitly said to just go (e.g. \"research X and report back\").\n\n### 3. Dispatch\n\nCluster related questions into research agents — don't spawn one agent per question, that's\nusually too fine-grained; don't dump every question on one agent either, that's too coarse to\nparallelize. A handful of clusters (roughly 3-8, scaled to topic breadth) is typical.\n\nLaunch them as **parallel forks** (`subagent_type: \"fork\"`) in a single message — per this user's\ntoken-discipline preference, forking beats fresh agents here because the raw search\u002Ffetch output\nisn't needed afterward, only the synthesized findings, and forks share this session's prompt\ncache. Each fork's prompt should:\n- Name its specific question cluster, not the whole topic (a fresh reader of just that prompt\n  should know exactly what to go find).\n- Be told to use WebSearch\u002FWebFetch and any other relevant tools, prefer primary sources over\n  secondhand summaries, and cite what it used (URLs, doc titles).\n- Be told to flag disagreement between sources and areas where it couldn't find a confident\n  answer, rather than smoothing over gaps.\n\nDo not race ahead and narrate or guess at findings before the forks report back — their results\narrive as later notifications, not something to predict.\n\n### 4. Synthesize\n\nOnce all forks report, merge their findings into one coherent report, organized by the same\nthemes as the question set (not by which agent answered what). For each theme: the answer(s), the\nconfidence level, and where sources disagree or coverage is thin. Note the actual sources used;\ndon't present findings as settled fact if only one weak source backs them.\n\nIf the report is long enough that the user would want to reference, share, or come back to it\nlater, offer it as an Artifact in one line rather than only dumping it into chat — don't publish\nwithout asking.\n\n### 5. Stop\n\nEnd by handing over the findings report. Do not automatically start `\u002Fgrill-me`, enter plan mode,\nor propose next steps as if they're already decided — the user reads the research and decides\nwhat, if anything, to interview themselves about next. It's fine to mention `\u002Fgrill-me` is there\nif they want to sharpen a resulting plan, but that's a pointer, not a continuation.\n","# Research\n\nTurn a loose topic into a synthesized, sourced findings report. This is the step *before*\n`\u002Fgrill-me` and plan mode, not a replacement for them — it ends by handing the user a report,\nnever by auto-continuing into an interview or a plan.\n\n## Process\n\n### 1. Discuss\n\nGet the topic and the reason it matters in a sentence or two: what decision or question this\nresearch is meant to inform. If the user already stated it, don't re-ask — just confirm you have\nenough to scope the question set. If it's genuinely vague (no worked-out angle at all), ask one\nclarifying question rather than guessing at scope; don't turn this into a `\u002Fgrill-me` interview.\n\n### 2. Draft the question set\n\nWrite an exhaustive set of sub-questions covering the topic, grouped by theme (e.g. prior art \u002F\nexisting solutions, technical tradeoffs, cost\u002Fpricing, risks and failure modes, ecosystem\u002Fvendor\nlandscape, what would change the answer). Aim for coverage, not padding — skip themes that don't\napply to this topic rather than forcing a fixed template.\n\nShow the numbered, grouped list to the user before dispatching anything. This is a cheap\ncheckpoint: dispatching a dozen parallel agents on the wrong question set is the expensive\nmistake it prevents. Let the user add, cut, or re-scope questions. Skip the pause only if the user\nexplicitly said to just go (e.g. \"research X and report back\").\n\n### 3. Dispatch\n\nCluster related questions into research agents — don't spawn one agent per question, that's\nusually too fine-grained; don't dump every question on one agent either, that's too coarse to\nparallelize. A handful of clusters (roughly 3-8, scaled to topic breadth) is typical.\n\nLaunch them as **parallel forks** (`subagent_type: \"fork\"`) in a single message — per this user's\ntoken-discipline preference, forking beats fresh agents here because the raw search\u002Ffetch output\nisn't needed afterward, only the synthesized findings, and forks share this session's prompt\ncache. Each fork's prompt should:\n- Name its specific question cluster, not the whole topic (a fresh reader of just that prompt\n  should know exactly what to go find).\n- Be told to use WebSearch\u002FWebFetch and any other relevant tools, prefer primary sources over\n  secondhand summaries, and cite what it used (URLs, doc titles).\n- Be told to flag disagreement between sources and areas where it couldn't find a confident\n  answer, rather than smoothing over gaps.\n\nDo not race ahead and narrate or guess at findings before the forks report back — their results\narrive as later notifications, not something to predict.\n\n### 4. Synthesize\n\nOnce all forks report, merge their findings into one coherent report, organized by the same\nthemes as the question set (not by which agent answered what). For each theme: the answer(s), the\nconfidence level, and where sources disagree or coverage is thin. Note the actual sources used;\ndon't present findings as settled fact if only one weak source backs them.\n\nIf the report is long enough that the user would want to reference, share, or come back to it\nlater, offer it as an Artifact in one line rather than only dumping it into chat — don't publish\nwithout asking.\n\n### 5. Stop\n\nEnd by handing over the findings report. Do not automatically start `\u002Fgrill-me`, enter plan mode,\nor propose next steps as if they're already decided — the user reads the research and decides\nwhat, if anything, to interview themselves about next. It's fine to mention `\u002Fgrill-me` is there\nif they want to sharpen a resulting plan, but that's a pointer, not a continuation.\n",{"name":5,"description":7,"disable-model-invocation":19},"master","df6f0f83d05e7ba7bee5e76189c60dd5834a8173","skills\u002Fresearch","https:\u002F\u002Fgithub.com\u002Fjpol34\u002Fpolyak-dev-kit","https:\u002F\u002Fgithub.com\u002Fjpol34","2026-09-18 03:22:58",{"source":8,"skillName":5,"commands":30,"prompt":40,"zipAvailable":41},[31,34,37],{"tool":32,"command":33},"npx","npx skills add jpol34\u002Fpolyak-dev-kit --skill research",{"tool":35,"command":36},"bunx","bunx skills add jpol34\u002Fpolyak-dev-kit --skill research",{"tool":38,"command":39},"pnpm","pnpm dlx skills add jpol34\u002Fpolyak-dev-kit --skill research","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fjpol34\u002Fpolyak-dev-kit\u002Fblob\u002Fmaster\u002Fskills\u002Fresearch\u002FSKILL.md",true,[],[44,51,58,65],{"skillNo":45,"name":46,"nameZh":6,"description":47,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":48,"categories":49,"keywords":50,"publishedAt":18},"01M7CYXCKJP92GP4TG2WTWETQ6","to-tickets","Break a plan, spec, or the current conversation into a set of tracer-bullet tickets, each declaring its blocking edges, published to the configured tracker (edges as text in one file per ticket locally, or native blocking links on a real tracker).","https:\u002F\u002Fgithub.com\u002Fjpol34\u002Fpolyak-dev-kit\u002Fblob\u002Fmaster\u002Fskills\u002Fto-tickets\u002FSKILL.md",[],[],{"skillNo":52,"name":53,"nameZh":6,"description":54,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":55,"categories":56,"keywords":57,"publishedAt":18},"01M7CYX5Q5Z4PHE4QYZJKH7PWC","grilling","Grill the user relentlessly about a plan, decision, or idea. Use when the user wants to stress-test their thinking, or uses any 'grill' trigger phrases.","https:\u002F\u002Fgithub.com\u002Fjpol34\u002Fpolyak-dev-kit\u002Fblob\u002Fmaster\u002Fskills\u002Fgrilling\u002FSKILL.md",[],[],{"skillNo":59,"name":60,"nameZh":6,"description":61,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":62,"categories":63,"keywords":64,"publishedAt":18},"01M7CYX49PT6SPAQCVK58WBF94","grill","A relentless interview to sharpen a plan or design.","https:\u002F\u002Fgithub.com\u002Fjpol34\u002Fpolyak-dev-kit\u002Fblob\u002Fmaster\u002Fskills\u002Fgrill\u002FSKILL.md",[],[],{"skillNo":66,"name":67,"nameZh":6,"description":68,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":69,"categories":70,"keywords":71,"publishedAt":18},"01M7CYWZPFW21ZVCH0VDCDW7FM","frontend","Detect a repo's current frontend state and bring it to a professional, cohesive look with minimal design decisions from the user — install or wire an appropriate component system, write a project DESIGN.md, and verify the result with a screenshot-critique loop. Manually invoked via \u002Ffrontend against any repo, regardless of how much frontend work already exists there.","https:\u002F\u002Fgithub.com\u002Fjpol34\u002Fpolyak-dev-kit\u002Fblob\u002Fmaster\u002Fskills\u002Ffrontend\u002FSKILL.md",[],[]]