[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7CQJY2VYTRT0BQ2V768TYHS":3,"skill-same:01M7CQJY2VYTRT0BQ2V768TYHS":59},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"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":25,"commitSha":26,"skillDirectory":27,"forksCount":13,"repositoryHtmlUrl":28,"repositoryHomepageUrl":6,"ownerHtmlUrl":29,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":30,"install":31,"resources":44},"01M7CQJY2VYTRT0BQ2V768TYHS","architecture-design",null,"Design and ratify ARCHITECTURE.md with the user: module boundaries, tech choices, folder layout, CI stages, then guard tests. Use to design or change a project's architecture. 架構設計 \u002F 架構規則 \u002F アーキテクチャ設計.","kouko\u002Floom-plugins","kouko","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F1480254?v=4","01M7CQCPAJR9C92XDGR0QDCQ98","01M7CQCPB8FTVY1BT26ENK1YP4",0,"MIT","https:\u002F\u002Fgithub.com\u002Fkouko\u002Floom-plugins\u002Fblob\u002Fmain\u002Floom-design\u002Fskills\u002Farchitecture-design\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: architecture-design\ndescription: |\n  Design and ratify ARCHITECTURE.md with the user: module boundaries, tech choices, folder layout, CI stages, then guard tests. Use to design or change a project's architecture. 架構設計 \u002F 架構規則 \u002F アーキテクチャ設計.\nversion: 1.0.0\n---\n\n## What this tool does\n\nRelative paths in this document are relative to this skill's own\ndirectory.\n\nDeliver local documents with actual absolute paths from the active worktree or\nexternal location, including in copyable instructions. For known local Markdown\nsupport, use the document's project-relative path as link text (external files:\nbasename) and absolute targets; otherwise show full absolute paths. Enclose\ntargets with spaces in angle brackets: `[doc.md](\u003C\u002Fworktree name\u002Fdoc.md>)`. Preserve bundled references, stored\nrelative paths, quoted examples and incidental filenames.\n\nThe user asks for an architecture for a repository. You design it with the\nuser — read what exists, propose options with trade-offs, let the user pick\n— then draft `ARCHITECTURE.md` in a temporary copy holding **the\ndecisions and the rules only**, write a guard test for every rule that can\nbe checked mechanically, restate them, and — on \"yes\" — write the\n`ratified-by:` line and install the ratified document and guards.\nFor a shared editable diagram during discussion, use\n`loom-workflow:loom-interactive-diagram`.\n\n## Step 1 — When to run\n\nWhen architecture boundaries or tradeoffs depend on missing consequential\nevidence, identify the decision and use `loom-design:design-research` as needed.\nReuse adequate evidence directly; return findings or uncertainty to this task\nwhile preserving user choices, scope and ratification authority.\n\nRun this **on request** only. `ARCHITECTURE.md` is **never required**: an\nabsent or unratified `ARCHITECTURE.md` never blocks a change, at any\nstation.\n\nWhen `ARCHITECTURE.md` already exists, run in **re-design mode**. Keep the\nratified root document and its active guards unchanged while drafting in a\ntemporary copy. When a\nchange alters the structure, re-design only the affected part with the\nuser (Step 2, for the affected choices alone), then update the affected\ndecisions, rules and guards in the **same commit**, run the validator, and\nre-ratify (Step 4) — a decision, rule or guard changed without the others\nleaves the document and the suite disagreeing.\n\n## Step 2 — Design with the user → ARCHITECTURE.md\n\n**Read first.** Read the project's requirements — its intent files,\nREADME and `PRINCIPLES.md`, when present — and the existing code: its\nlayout, languages, build files, test setup and CI, if any.\n\n**Propose.** The open choices are:\n\n1. **Module split and dependency direction** — which modules exist, which\n   may depend on which, and which never.\n2. **Main technology choices** — language, framework, key libraries, where\n   the existing code has not already settled them.\n3. **Folder structure** — where each kind of file goes (code, tests,\n   scripts), and the size a file stays under.\n4. **Required CI stages** — what CI must run, and in what order.\n\nA choice the existing code already settles is stated, not re-opened —\nunless the user asks to revisit it. For each open choice, present **at\nleast two options** in MADR style: the decision drivers first, then each\noption's trade-off against those drivers (\"good, because … \u002F bad,\nbecause …\"), then your recommendation and what would make it wrong.\n**Never present a single answer**, even when one option looks obvious.\nDraw the options from `references\u002Fdesign-know-how.md`. The user picks;\non a question, answer it and present the options again.\n\nData models and API interfaces stay out of this design: they belong to\neach change's spec.\n\n**Write.** Work in a temporary copy of the repository outside its normal\ntest discovery paths. Write the draft `ARCHITECTURE.md` at that copy's root,\nwithout a `ratified-by:` line, following\n`references\u002Farchitecture-md-schema.md`: a title, a `## Decisions` section\nrecording each pick (the choice, the options considered, the reason), then\nexactly the four rule sections, one line per rule, and no overview or\nbackground section.\n\n## Step 3 — Write the guards and validate\n\nIn the temporary copy, for every rule that can be checked mechanically,\nwrite a guard test in the repository's **own test framework**, under a path its `package-tests:`\ncommand in `docs\u002Floom\u002FKICKOFF-DEFAULTS.md` already runs — or extend that\ncommand so it does; a guard the suite never runs holds nothing. When that\nfile or its `package-tests:` line is absent or `none`, propose the\nrepository's test command to the user and record it there as\n`- package-tests: \u003Ccommand> — \u003Creason> (\u003Cdate>)` before writing guards.\nRecord the guard's path on the rule line (`check: \u003Cguard path>`). A rule\nthat needs judgment says `check: review` and gets no guard.\n\nEach guard's failure message follows the schema's \"Guard failure message\"\nsection. Run the guards once in the temporary copy; a guard that fails on today's code is either a rule the user must\nrestate or a violation to show them.\n\nThen check the file, with `\u003Cloom-design>` standing for this plugin's own\ncheckout:\n\n```\npython3 \u003Cloom-design>\u002Fscripts\u002Farchitecture-design\u002Fvalidate_architecture_output.py ARCHITECTURE.md --draft\n```\n\nExit 0: go to Step 4. Non-zero: fix what it names and run it again. Never\nread back a file that has not exited 0 here.\n\n## Step 4 — Restate and ratify\n\nRead every decision and rule back to the user in plain words, with which\nrules a guard checks and which a reviewer judges. On a correction, fix it\nand read back again. On \"yes\" — and only then — write one line directly under the title:\n\n```\nratified-by: \u003Cname> \u003Cdate>\n```\n\n`\u003Cname>` is the user's own name or handle; `\u003Cdate>` is today, `YYYY-MM-DD`.\nRun the validator again without `--draft` in the temporary copy. On success,\ninstall the ratified document, guards and associated suite configuration in\nthe working repository together, replacing the old versions in re-design\nmode. Activate the guards only after ratification. Run the normal suite;\nnever commit an unvalidated draft as the active architecture.\n\n## Step 5 — Commit\n\nAdd `ARCHITECTURE.md`, the guard files, and every file Step 3 edited so the\nsuite runs the guards (`docs\u002Floom\u002FKICKOFF-DEFAULTS.md`, test or tool\nconfig) by name, then:\n\n```\ngit commit -m \"docs(loom): ARCHITECTURE.md ratified\"\n```\n\n## Downstream — how the rest of the flow uses it\n\n`write-plan` reads `ARCHITECTURE.md`, when ratified with the existing\n`ratified-by: \u003Cname> \u003Cdate>` line, before its Task DAG: it\nplaces added or moved files by the rules and names the rule id on the\ntask's Risk line. At `closing-review`, the code lens dimension\n`architecture-conformance` scores the diff against the rules — a violation\nis a finding with a fix; with no ratified `ARCHITECTURE.md` the dimension is scored\n**N\u002FA with a one-line reason**. A missing `ARCHITECTURE.md` is only the\nstanding WARN the checker prints and never blocks any station.\n","## What this tool does\n\nRelative paths in this document are relative to this skill's own\ndirectory.\n\nDeliver local documents with actual absolute paths from the active worktree or\nexternal location, including in copyable instructions. For known local Markdown\nsupport, use the document's project-relative path as link text (external files:\nbasename) and absolute targets; otherwise show full absolute paths. Enclose\ntargets with spaces in angle brackets: `[doc.md](\u003C\u002Fworktree name\u002Fdoc.md>)`. Preserve bundled references, stored\nrelative paths, quoted examples and incidental filenames.\n\nThe user asks for an architecture for a repository. You design it with the\nuser — read what exists, propose options with trade-offs, let the user pick\n— then draft `ARCHITECTURE.md` in a temporary copy holding **the\ndecisions and the rules only**, write a guard test for every rule that can\nbe checked mechanically, restate them, and — on \"yes\" — write the\n`ratified-by:` line and install the ratified document and guards.\nFor a shared editable diagram during discussion, use\n`loom-workflow:loom-interactive-diagram`.\n\n## Step 1 — When to run\n\nWhen architecture boundaries or tradeoffs depend on missing consequential\nevidence, identify the decision and use `loom-design:design-research` as needed.\nReuse adequate evidence directly; return findings or uncertainty to this task\nwhile preserving user choices, scope and ratification authority.\n\nRun this **on request** only. `ARCHITECTURE.md` is **never required**: an\nabsent or unratified `ARCHITECTURE.md` never blocks a change, at any\nstation.\n\nWhen `ARCHITECTURE.md` already exists, run in **re-design mode**. Keep the\nratified root document and its active guards unchanged while drafting in a\ntemporary copy. When a\nchange alters the structure, re-design only the affected part with the\nuser (Step 2, for the affected choices alone), then update the affected\ndecisions, rules and guards in the **same commit**, run the validator, and\nre-ratify (Step 4) — a decision, rule or guard changed without the others\nleaves the document and the suite disagreeing.\n\n## Step 2 — Design with the user → ARCHITECTURE.md\n\n**Read first.** Read the project's requirements — its intent files,\nREADME and `PRINCIPLES.md`, when present — and the existing code: its\nlayout, languages, build files, test setup and CI, if any.\n\n**Propose.** The open choices are:\n\n1. **Module split and dependency direction** — which modules exist, which\n   may depend on which, and which never.\n2. **Main technology choices** — language, framework, key libraries, where\n   the existing code has not already settled them.\n3. **Folder structure** — where each kind of file goes (code, tests,\n   scripts), and the size a file stays under.\n4. **Required CI stages** — what CI must run, and in what order.\n\nA choice the existing code already settles is stated, not re-opened —\nunless the user asks to revisit it. For each open choice, present **at\nleast two options** in MADR style: the decision drivers first, then each\noption's trade-off against those drivers (\"good, because … \u002F bad,\nbecause …\"), then your recommendation and what would make it wrong.\n**Never present a single answer**, even when one option looks obvious.\nDraw the options from `references\u002Fdesign-know-how.md`. The user picks;\non a question, answer it and present the options again.\n\nData models and API interfaces stay out of this design: they belong to\neach change's spec.\n\n**Write.** Work in a temporary copy of the repository outside its normal\ntest discovery paths. Write the draft `ARCHITECTURE.md` at that copy's root,\nwithout a `ratified-by:` line, following\n`references\u002Farchitecture-md-schema.md`: a title, a `## Decisions` section\nrecording each pick (the choice, the options considered, the reason), then\nexactly the four rule sections, one line per rule, and no overview or\nbackground section.\n\n## Step 3 — Write the guards and validate\n\nIn the temporary copy, for every rule that can be checked mechanically,\nwrite a guard test in the repository's **own test framework**, under a path its `package-tests:`\ncommand in `docs\u002Floom\u002FKICKOFF-DEFAULTS.md` already runs — or extend that\ncommand so it does; a guard the suite never runs holds nothing. When that\nfile or its `package-tests:` line is absent or `none`, propose the\nrepository's test command to the user and record it there as\n`- package-tests: \u003Ccommand> — \u003Creason> (\u003Cdate>)` before writing guards.\nRecord the guard's path on the rule line (`check: \u003Cguard path>`). A rule\nthat needs judgment says `check: review` and gets no guard.\n\nEach guard's failure message follows the schema's \"Guard failure message\"\nsection. Run the guards once in the temporary copy; a guard that fails on today's code is either a rule the user must\nrestate or a violation to show them.\n\nThen check the file, with `\u003Cloom-design>` standing for this plugin's own\ncheckout:\n\n```\npython3 \u003Cloom-design>\u002Fscripts\u002Farchitecture-design\u002Fvalidate_architecture_output.py ARCHITECTURE.md --draft\n```\n\nExit 0: go to Step 4. Non-zero: fix what it names and run it again. Never\nread back a file that has not exited 0 here.\n\n## Step 4 — Restate and ratify\n\nRead every decision and rule back to the user in plain words, with which\nrules a guard checks and which a reviewer judges. On a correction, fix it\nand read back again. On \"yes\" — and only then — write one line directly under the title:\n\n```\nratified-by: \u003Cname> \u003Cdate>\n```\n\n`\u003Cname>` is the user's own name or handle; `\u003Cdate>` is today, `YYYY-MM-DD`.\nRun the validator again without `--draft` in the temporary copy. On success,\ninstall the ratified document, guards and associated suite configuration in\nthe working repository together, replacing the old versions in re-design\nmode. Activate the guards only after ratification. Run the normal suite;\nnever commit an unvalidated draft as the active architecture.\n\n## Step 5 — Commit\n\nAdd `ARCHITECTURE.md`, the guard files, and every file Step 3 edited so the\nsuite runs the guards (`docs\u002Floom\u002FKICKOFF-DEFAULTS.md`, test or tool\nconfig) by name, then:\n\n```\ngit commit -m \"docs(loom): ARCHITECTURE.md ratified\"\n```\n\n## Downstream — how the rest of the flow uses it\n\n`write-plan` reads `ARCHITECTURE.md`, when ratified with the existing\n`ratified-by: \u003Cname> \u003Cdate>` line, before its Task DAG: it\nplaces added or moved files by the rules and names the rule id on the\ntask's Risk line. At `closing-review`, the code lens dimension\n`architecture-conformance` scores the diff against the rules — a violation\nis a finding with a fix; with no ratified `ARCHITECTURE.md` the dimension is scored\n**N\u002FA with a one-line reason**. A missing `ARCHITECTURE.md` is only the\nstanding WARN the checker prints and never blocks any station.\n",{"name":5,"version":23,"description":24},"1.0.0","Design and ratify ARCHITECTURE.md with the user: module boundaries, tech choices, folder layout, CI stages, then guard tests. Use to design or change a project's architecture. 架構設計 \u002F 架構規則 \u002F アーキテクチャ設計.\n","main","3b1a9972644517911454530dc1876f88525b1f37","loom-design\u002Fskills\u002Farchitecture-design","https:\u002F\u002Fgithub.com\u002Fkouko\u002Floom-plugins","https:\u002F\u002Fgithub.com\u002Fkouko","2026-11-15 00:45:02",{"source":8,"skillName":5,"commands":32,"prompt":42,"zipAvailable":43},[33,36,39],{"tool":34,"command":35},"npx","npx skills add kouko\u002Floom-plugins --skill architecture-design",{"tool":37,"command":38},"bunx","bunx skills add kouko\u002Floom-plugins --skill architecture-design",{"tool":40,"command":41},"pnpm","pnpm dlx skills add kouko\u002Floom-plugins --skill architecture-design","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fkouko\u002Floom-plugins\u002Fblob\u002Fmain\u002Floom-design\u002Fskills\u002Farchitecture-design\u002FSKILL.md",true,[45,51,55],{"path":46,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":49,"sourceHtmlUrl":50},"README.md","REFERENCE","参考文档",1891,"https:\u002F\u002Fgithub.com\u002Fkouko\u002Floom-plugins\u002Fblob\u002Fmain\u002Floom-design\u002Fskills\u002Farchitecture-design\u002FREADME.md",{"path":52,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":53,"sourceHtmlUrl":54},"references\u002Farchitecture-md-schema.md",3546,"https:\u002F\u002Fgithub.com\u002Fkouko\u002Floom-plugins\u002Fblob\u002Fmain\u002Floom-design\u002Fskills\u002Farchitecture-design\u002Freferences\u002Farchitecture-md-schema.md",{"path":56,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":57,"sourceHtmlUrl":58},"references\u002Fdesign-know-how.md",8929,"https:\u002F\u002Fgithub.com\u002Fkouko\u002Floom-plugins\u002Fblob\u002Fmain\u002Floom-design\u002Fskills\u002Farchitecture-design\u002Freferences\u002Fdesign-know-how.md",[60,67,74,81,88,95],{"skillNo":61,"name":62,"nameZh":6,"description":63,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":64,"categories":65,"keywords":66,"publishedAt":18},"01M7CQNVD2W5MP7PKX3MD0YSQE","using-loom-workflow","Route broad Loom workflow-tool requests to the right skill. Use when the request spans tools or the right workflow skill is unclear.","https:\u002F\u002Fgithub.com\u002Fkouko\u002Floom-plugins\u002Fblob\u002Fmain\u002Floom-workflow\u002Fskills\u002Fusing-loom-workflow\u002FSKILL.md",[],[],{"skillNo":68,"name":69,"nameZh":6,"description":70,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":71,"categories":72,"keywords":73,"publishedAt":18},"01M7CQNNTPVG52CS2C6Q9HJ2E2","skill-retro","Review selected past or current Claude Code or Codex sessions against readable target skill sources and related integration. Produce source-linked proposals without applying changes.","https:\u002F\u002Fgithub.com\u002Fkouko\u002Floom-plugins\u002Fblob\u002Fmain\u002Floom-workflow\u002Fskills\u002Fskill-retro\u002FSKILL.md",[],[],{"skillNo":75,"name":76,"nameZh":6,"description":77,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":78,"categories":79,"keywords":80,"publishedAt":18},"01M7CQNHR1BDKJ63CC4R1RY9GE","recap-state","Give an in-session recap of current progress for 'where were we', '我們剛剛在幹嘛', '剛剛講到哪', '我跟丟了', 'ちょっと振り返って', '今どこだっけ', or '振り返り'. The built-in \u002Frecap is an away-summary.","https:\u002F\u002Fgithub.com\u002Fkouko\u002Floom-plugins\u002Fblob\u002Fmain\u002Floom-workflow\u002Fskills\u002Frecap-state\u002FSKILL.md",[],[],{"skillNo":82,"name":83,"nameZh":6,"description":84,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":85,"categories":86,"keywords":87,"publishedAt":18},"01M7CQNE1A9272WPRE1MM3JRFK","project-retro","Review one current or specified coding session for source-backed project\ndevelopment improvements. Use for a requested retrospective, including\n振り返り seeking improvements; not for current-progress recaps or recurring skill mining.","https:\u002F\u002Fgithub.com\u002Fkouko\u002Floom-plugins\u002Fblob\u002Fmain\u002Floom-workflow\u002Fskills\u002Fproject-retro\u002FSKILL.md",[],[],{"skillNo":89,"name":90,"nameZh":6,"description":91,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":92,"categories":93,"keywords":94,"publishedAt":18},"01M7CQNACJB36T0A039HT89P38","loom-visualization","Show comparisons, flows, decisions, states or reasoning chains as tables, ASCII or Mermaid in coding chat, including when a Loom station reports to the user (intent restatement, choices, acceptance test results); not Obsidian notes.","https:\u002F\u002Fgithub.com\u002Fkouko\u002Floom-plugins\u002Fblob\u002Fmain\u002Floom-workflow\u002Fskills\u002Floom-visualization\u002FSKILL.md",[],[],{"skillNo":96,"name":97,"nameZh":6,"description":98,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":99,"categories":100,"keywords":101,"publishedAt":18},"01M7CQN7TVS0XF44JXQBWA32Z1","loom-memory","Record repo-wide lessons when saving to your own memory or asked to remember; recall, reconcile, or forget them when a task needs it.","https:\u002F\u002Fgithub.com\u002Fkouko\u002Floom-plugins\u002Fblob\u002Fmain\u002Floom-workflow\u002Fskills\u002Floom-memory\u002FSKILL.md",[],[]]