[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M69BWB3PEQWAMWFJM8H4GSNX":3,"skill-same:01M69BWB3PEQWAMWFJM8H4GSNX":79},{"skillNo":4,"name":5,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":21,"frontMatter":22,"refName":23,"commitSha":24,"skillDirectory":25,"forksCount":26,"repositoryHtmlUrl":27,"repositoryHomepageUrl":6,"ownerHtmlUrl":28,"ownerBlogUrl":29,"ownerDescription":30,"updatedAt":31,"install":32,"resources":45},"01M69BWB3PEQWAMWFJM8H4GSNX","add_smart_capability",null,"microsoft\u002Famplifier-smart-tool-creator","microsoft","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F6154722?v=4","01M698VYDXZ6JW5WH3CRV650R8","Microsoft","01M69BW701FWQCA7V57HG6B4BF",1,"MIT","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Famplifier-smart-tool-creator\u002Fblob\u002Fmain\u002Fsrc\u002Fsmart_tool_creator\u002Fcapabilities\u002Fadd_smart_capability\u002FSKILL.md",[],[18],"add smart capability","2026-11-10 10:12:30",false,"`add-smart-capability` extends a smart tool that already exists: an agent reads the tool,\nimplements one model-backed capability in its library, exposes it from the CLI, writes the\ntests and the docs, then runs the tool's own checks (`uv run pytest`, the conformance kit,\nand `prek run --all-files` when `prek` is installed) and fixes what they report. Nothing is\ncommitted; the working tree is left for you to review.\n\n```bash\nsmart-tool-creator add-smart-capability \\\n  \"Given an incident id, fetch its chat transcript and alert timeline from the incident platform and draft a blameless postmortem: summary, impact, contributing factors, and action items with owners\" \\\n  --directory ~\u002Fsrc\u002Fincident-postmortem \\\n  --context \"The platform client is src\u002Fincident_postmortem\u002Fplatform.py; fetch through it, never call the API directly\" \\\n  --context \"Our postmortem template is at ~\u002Fnotes\u002Fpostmortem-template.md; match its headings\"\n```\n\n```python\nfrom pathlib import Path\n\nfrom smart_tool_creator.lib import add_smart_capability\n\nadded = add_smart_capability(\n    \"Given an incident id, fetch its chat transcript and alert timeline from the incident platform \"\n    \"and draft a blameless postmortem: summary, impact, contributing factors, and action items with owners\",\n    directory=Path(\"~\u002Fsrc\u002Fincident-postmortem\").expanduser(),\n    context=[\n        \"The platform client is src\u002Fincident_postmortem\u002Fplatform.py; fetch through it, never call the API directly\",\n        \"Our postmortem template is at ~\u002Fnotes\u002Fpostmortem-template.md; match its headings\",\n    ],\n)\nadded.report, added.checks, added.fix_rounds, added.output_message\n```\n\n## Arguments\n\n- `REQUEST`: the whole brief for one capability: what it does, for whom, and what it takes in\n  and gives back. Required.\n- `--directory PATH`: the smart tool to work in; the current directory when omitted. It must\n  hold a `smart-tool.json` at its root, and no parent is searched, so a workspace holding\n  several tools is never extended by accident. It must also have an `Intelligence` interface\n  at `src\u002F\u003Cpackage>\u002Fintelligence\u002Finterface.py` for the capability to run through, which a\n  tool scaffolded with `--no-agent-providers` does not.\n- `--context TEXT`: repeatable free text, usually paths to notes, transcripts, or exemplars\n  the agent should read before it designs anything. It reads the paths itself, so name them\n  rather than pasting their contents.\n- `--agent-provider`: what the agent runs through, `{{ agent_providers | join: \"` or `\" }}`. The\n  first installed, in that order, when omitted.\n- `--model`: the model the agent runs on: a Copilot model id for `copilot`,\n  `\u003Cprovider>\u002F\u003Cmodel>` for `amplifier-agent` (for instance\n  `{{ default_intelligence_models[\"amplifier-agent\"] }}`), a Codex model id for `codex`, a\n  Claude model id for `claude`. Defaults to `{{ default_intelligence_models.copilot }}` on\n  `copilot`, `{{ default_intelligence_models[\"amplifier-agent\"] }}` on `amplifier-agent`,\n  `{{ default_intelligence_models.codex }}` on `codex`, and\n  `{{ default_intelligence_models.claude }}` on `claude`.\n- `--reasoning-effort`: how hard the model thinks before it acts, one of `low`, `medium`,\n  `high`, `xhigh`, `max`. Defaults to `{{ default_intelligence_reasoning_effort }}`. Applies to\n  the `copilot`, `codex`, and `claude` agent providers.\n- `intelligence`, library only: the `Intelligence` implementation the agent runs through,\n  which wins over `agent_provider`; `resolve_intelligence(agent_provider)` when omitted. Tests\n  inject a fake.\n\n## Result\n\n`AddedCapability` carries `root`, the extended tool's distribution root; `report`, the\nagent's final message, which names the capability, the files it touched, the command to try\nit, and its caveats; `checks`, one entry per check with its `name`, `command`, `status`\n(`passed`, `failed`, or `skipped`), and the tail of its output when it failed or the reason\nwhen it was skipped; `fix_rounds`, the extra agent runs spent on failing checks, at most 2;\nand `output_message`, all of that for the calling agent, which is what the CLI prints.\n\nThe prek check is skipped, never failed, when the tool has no `.pre-commit-config.yaml` or\n`prek` is not on `PATH`; the conformance check is skipped when the kit itself could not run,\nand names the failing rules when it fails. A check still failing when the work stops is returned rather than\nraised and named in the message, and the CLI exits 1; the work stays in the tool's working\ntree either way, and what it is worth is the caller's call.\n",{},"main","95536af8d4c33d2d6ef346fa8c2759a1b1c5d9c6","src\u002Fsmart_tool_creator\u002Fcapabilities\u002Fadd_smart_capability",0,"https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Famplifier-smart-tool-creator","https:\u002F\u002Fgithub.com\u002Fmicrosoft","https:\u002F\u002Fopensource.microsoft.com","Open source projects and samples from Microsoft","2026-10-30 07:24:08",{"source":7,"skillName":5,"commands":33,"prompt":43,"zipAvailable":44},[34,37,40],{"tool":35,"command":36},"npx","npx skills add microsoft\u002Famplifier-smart-tool-creator --skill add_smart_capability",{"tool":38,"command":39},"bunx","bunx skills add microsoft\u002Famplifier-smart-tool-creator --skill add_smart_capability",{"tool":41,"command":42},"pnpm","pnpm dlx skills add microsoft\u002Famplifier-smart-tool-creator --skill add_smart_capability","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Famplifier-smart-tool-creator\u002Fblob\u002Fmain\u002Fsrc\u002Fsmart_tool_creator\u002Fcapabilities\u002Fadd_smart_capability\u002FSKILL.md",true,[46,51,57,61,65,69,75],{"path":47,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":26,"sourceHtmlUrl":50},"__init__.py","SCRIPT","脚本","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Famplifier-smart-tool-creator\u002Fblob\u002Fmain\u002Fsrc\u002Fsmart_tool_creator\u002Fcapabilities\u002Fadd_smart_capability\u002F__init__.py",{"path":52,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":55,"sourceHtmlUrl":56},"add_smart_capability.md.liquid","OTHER","其他",6142,"https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Famplifier-smart-tool-creator\u002Fblob\u002Fmain\u002Fsrc\u002Fsmart_tool_creator\u002Fcapabilities\u002Fadd_smart_capability\u002Fadd_smart_capability.md.liquid",{"path":58,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":59,"sourceHtmlUrl":60},"capability.py",8986,"https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Famplifier-smart-tool-creator\u002Fblob\u002Fmain\u002Fsrc\u002Fsmart_tool_creator\u002Fcapabilities\u002Fadd_smart_capability\u002Fcapability.py",{"path":62,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":63,"sourceHtmlUrl":64},"fix_checks.md.liquid",900,"https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Famplifier-smart-tool-creator\u002Fblob\u002Fmain\u002Fsrc\u002Fsmart_tool_creator\u002Fcapabilities\u002Fadd_smart_capability\u002Ffix_checks.md.liquid",{"path":66,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":67,"sourceHtmlUrl":68},"output_message.md.liquid",703,"https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Famplifier-smart-tool-creator\u002Fblob\u002Fmain\u002Fsrc\u002Fsmart_tool_creator\u002Fcapabilities\u002Fadd_smart_capability\u002Foutput_message.md.liquid",{"path":70,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":73,"sourceHtmlUrl":74},"references\u002Fcode-patterns.md","REFERENCE","参考文档",4587,"https:\u002F\u002Fgithub.com\u002Fgithub\u002Fawesome-copilot\u002Fblob\u002Fmain\u002Fskills\u002Fdotnet-timezone\u002Freferences\u002Fcode-patterns.md",{"path":76,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":77,"sourceHtmlUrl":78},"references\u002Ftimezone-index.md",4352,"https:\u002F\u002Fgithub.com\u002Fgithub\u002Fawesome-copilot\u002Fblob\u002Fmain\u002Fskills\u002Fdotnet-timezone\u002Freferences\u002Ftimezone-index.md",[80,86,94,101],{"skillNo":81,"name":82,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":83,"categories":84,"keywords":85,"publishedAt":19},"01M69BWGZ1HQKVQ8H0D7D877H3","init","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Famplifier-smart-tool-creator\u002Fblob\u002Fmain\u002Fsrc\u002Fsmart_tool_creator\u002Fcapabilities\u002Finit\u002FSKILL.md",[],[82],{"skillNo":87,"name":88,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":89,"categories":90,"keywords":91,"publishedAt":19},"01M69BWDYRDS233M26R595EZ01","check_spec_adherence","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Famplifier-smart-tool-creator\u002Fblob\u002Fmain\u002Fsrc\u002Fsmart_tool_creator\u002Fcapabilities\u002Fcheck_spec_adherence\u002FSKILL.md",[],[92,93],"spec","check spec adherence",{"skillNo":95,"name":96,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":97,"categories":98,"keywords":99,"publishedAt":19},"01M69BWCM24AVRW9VTYBZAHJQW","check_conformance","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Famplifier-smart-tool-creator\u002Fblob\u002Fmain\u002Fsrc\u002Fsmart_tool_creator\u002Fcapabilities\u002Fcheck_conformance\u002FSKILL.md",[],[100],"check conformance",{"skillNo":102,"name":103,"nameZh":6,"description":104,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":105,"categories":106,"keywords":107,"publishedAt":19},"01M69BW97VYEG1084VTZA2JRHB","smart-tool-creator","Create, validate, and evaluate smart tools that follow the Amplifier Smart Tool Spec, by driving the smart-tool-creator smart tool from a CLI, a library, or an agent. Use when (1) packaging domain expertise as a standalone tool any agent can call, (2) checking an existing smart tool against the spec and its conformance kit, (3) measuring how well a smart tool's model-backed capabilities work. Triggers on \"smart tool\", \"smart-tool-creator\", \"amplifier smart tool\", \"conformance kit\", \"SMART_TOOL.md\", \"smart-tool.json\".","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Famplifier-smart-tool-creator\u002Fblob\u002Fmain\u002Fskills\u002Fsmart-tool-creator\u002FSKILL.md",[],[108,109,92,110,8,111,112,113],"smart tool creator","cli","github","validate","create","capabilities"]