[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M758MH3AXGK2B44931MBV0AH":3,"skill-same:01M758MH3AXGK2B44931MBV0AH":49},{"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":24,"commitSha":25,"skillDirectory":26,"forksCount":13,"repositoryHtmlUrl":27,"repositoryHomepageUrl":6,"ownerHtmlUrl":28,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":29,"install":30,"resources":42},"01M758MH3AXGK2B44931MBV0AH","improve-codebase-architecture",null,"Scan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.","joeloverbeck\u002Fcontinuity-loom","joeloverbeck","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F120873?v=4","01M758KCBJ0GWAPGN74C9NJF8T","01M758KCCNQ8F0YS327T45F7C0",0,"GPL-3.0","https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fcontinuity-loom\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fimprove-codebase-architecture\u002FSKILL.md",[],[],"2026-11-12 13:55:43",false,"---\nname: improve-codebase-architecture\ndescription: Scan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.\ndisable-model-invocation: true\n---\n\n# Improve Codebase Architecture\n\nSurface architectural friction and propose **deepening opportunities** that\nimprove testability and AI-navigability by turning shallow modules into deep\nones.\n\nLoad `\u002Fcodebase-design` for the architecture vocabulary, principles, dependency\ncategories, and repository-authority rules. Use the relevant domain glossary's\nterms for the code under review, and do not re-litigate governing principles or\nADRs without concrete friction that warrants reopening them.\n\n## 1. Explore and verify\n\nFollow the repository entrypoint and the domain and architecture authorities\nidentified by `\u002Fcodebase-design`. Missing optional domain or ADR files are not a\nblocker.\n\nUse authorized Explore agents for breadth when the codebase warrants fan-out.\nWhen delegation is unavailable or policy-disallowed, scan locally and state that\nreason. In either case, do the depth and evidence work directly: read the\nrelevant files and call sites yourself, and support promoted candidates with\ncompact counts, deletion-test results, and exact line references.\n\nExplore organically:\n\n- Where does one concept require bouncing among many small modules?\n- Which modules are shallow?\n- Where were pure helpers extracted for tests while call-site bugs stayed hidden?\n- Where does coupled behavior leak across seams?\n- What is difficult to test through its current interface?\n\nApply `\u002Fcodebase-design`'s deletion test. Promote only friction that survives\ncall-site and authority verification; if a type, caller, test, principle, or ADR\nalready neutralizes the claim, drop or downgrade it.\n\n## 2. Present the HTML report\n\nWrite a fresh self-contained report outside the repository at\n`$TMPDIR\u002Farchitecture-review-\u003Ctimestamp>.html`, falling back to `\u002Ftmp` (or\n`%TEMP%` on Windows). Follow [HTML-REPORT.md](HTML-REPORT.md) for the visual,\nevidence, vocabulary, and card contract.\n\nOpen the report when the environment allows, using the OS-appropriate command;\non WSL prefer `wslview` or Windows Explorer. Dispatch terminal openers detached\nso they cannot hold the run open. When command execution is sandboxed, obtain\nthe environment's normal GUI or host-execution authorization before dispatching\nthe opener. Capture its immediate stderr and exit status; treat explicit\nlauncher refusal output, including `snap-confine` refusing to continue, as a\nfailed open even when `xdg-open` or `gio` returns zero. Apply the same GUI or\nhost authorization to any named-browser fallback instead of retrying it inside\nthe restricted sandbox. Always print both the exact opener command and the\nabsolute report path, because dispatch success does not prove a visible GUI.\nFailure to open a GUI does not invalidate a report that exists at the printed\npath.\n\nUse relevant principle or ADR callouts only when an authority conflicts with or\nendorses a verified candidate. End with one Top recommendation.\n\nDo **not** propose interfaces in the report. After writing it, ask exactly:\n\"Which of these would you like to explore?\"\n\n## 3. Grill the selected candidate\n\nOnce the user picks a candidate, run `\u002Fgrilling` to resolve constraints,\ndependencies, module shape, seam placement, hidden implementation, and surviving\ntests. Respect every explicit mutation limit from the user.\n\nUse `\u002Fdomain-modeling` when resolved terms or durable decisions require domain\nartifacts, and let that skill own glossary\u002FADR routing and closeout. For\nalternative interface designs, use `\u002Fcodebase-design`'s design-it-twice flow.\n\n## 4. Hand off authorized implementation\n\nIf the user asks to implement after grilling, first recap the selected modules,\ndomain and architectural authorities, accepted constraints, domain-model changes,\nbehavior posture, and expected tests\u002Fdocs. Default to behavior-preserving work\nunless the user explicitly approved a behavior change.\n\nThen use the repository's implementation workflow (`\u002Fimplement` for published\nissue or PRD work; otherwise the local coding guidelines). Keep edits scoped to\nthe agreed candidate and verify focused tests plus relevant full gates, including\ndoc, ADR, or principle checks when those authorities changed or justified the\ndesign.\n","# Improve Codebase Architecture\n\nSurface architectural friction and propose **deepening opportunities** that\nimprove testability and AI-navigability by turning shallow modules into deep\nones.\n\nLoad `\u002Fcodebase-design` for the architecture vocabulary, principles, dependency\ncategories, and repository-authority rules. Use the relevant domain glossary's\nterms for the code under review, and do not re-litigate governing principles or\nADRs without concrete friction that warrants reopening them.\n\n## 1. Explore and verify\n\nFollow the repository entrypoint and the domain and architecture authorities\nidentified by `\u002Fcodebase-design`. Missing optional domain or ADR files are not a\nblocker.\n\nUse authorized Explore agents for breadth when the codebase warrants fan-out.\nWhen delegation is unavailable or policy-disallowed, scan locally and state that\nreason. In either case, do the depth and evidence work directly: read the\nrelevant files and call sites yourself, and support promoted candidates with\ncompact counts, deletion-test results, and exact line references.\n\nExplore organically:\n\n- Where does one concept require bouncing among many small modules?\n- Which modules are shallow?\n- Where were pure helpers extracted for tests while call-site bugs stayed hidden?\n- Where does coupled behavior leak across seams?\n- What is difficult to test through its current interface?\n\nApply `\u002Fcodebase-design`'s deletion test. Promote only friction that survives\ncall-site and authority verification; if a type, caller, test, principle, or ADR\nalready neutralizes the claim, drop or downgrade it.\n\n## 2. Present the HTML report\n\nWrite a fresh self-contained report outside the repository at\n`$TMPDIR\u002Farchitecture-review-\u003Ctimestamp>.html`, falling back to `\u002Ftmp` (or\n`%TEMP%` on Windows). Follow [HTML-REPORT.md](HTML-REPORT.md) for the visual,\nevidence, vocabulary, and card contract.\n\nOpen the report when the environment allows, using the OS-appropriate command;\non WSL prefer `wslview` or Windows Explorer. Dispatch terminal openers detached\nso they cannot hold the run open. When command execution is sandboxed, obtain\nthe environment's normal GUI or host-execution authorization before dispatching\nthe opener. Capture its immediate stderr and exit status; treat explicit\nlauncher refusal output, including `snap-confine` refusing to continue, as a\nfailed open even when `xdg-open` or `gio` returns zero. Apply the same GUI or\nhost authorization to any named-browser fallback instead of retrying it inside\nthe restricted sandbox. Always print both the exact opener command and the\nabsolute report path, because dispatch success does not prove a visible GUI.\nFailure to open a GUI does not invalidate a report that exists at the printed\npath.\n\nUse relevant principle or ADR callouts only when an authority conflicts with or\nendorses a verified candidate. End with one Top recommendation.\n\nDo **not** propose interfaces in the report. After writing it, ask exactly:\n\"Which of these would you like to explore?\"\n\n## 3. Grill the selected candidate\n\nOnce the user picks a candidate, run `\u002Fgrilling` to resolve constraints,\ndependencies, module shape, seam placement, hidden implementation, and surviving\ntests. Respect every explicit mutation limit from the user.\n\nUse `\u002Fdomain-modeling` when resolved terms or durable decisions require domain\nartifacts, and let that skill own glossary\u002FADR routing and closeout. For\nalternative interface designs, use `\u002Fcodebase-design`'s design-it-twice flow.\n\n## 4. Hand off authorized implementation\n\nIf the user asks to implement after grilling, first recap the selected modules,\ndomain and architectural authorities, accepted constraints, domain-model changes,\nbehavior posture, and expected tests\u002Fdocs. Default to behavior-preserving work\nunless the user explicitly approved a behavior change.\n\nThen use the repository's implementation workflow (`\u002Fimplement` for published\nissue or PRD work; otherwise the local coding guidelines). Keep edits scoped to\nthe agreed candidate and verify focused tests plus relevant full gates, including\ndoc, ADR, or principle checks when those authorities changed or justified the\ndesign.\n",{"name":5,"description":7,"disable-model-invocation":23},true,"main","bb13c72ec4134361fec0b0584913aee821330c02",".claude\u002Fskills\u002Fimprove-codebase-architecture","https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fcontinuity-loom","https:\u002F\u002Fgithub.com\u002Fjoeloverbeck","2026-11-10 23:00:02",{"source":8,"skillName":5,"commands":31,"prompt":41,"zipAvailable":23},[32,35,38],{"tool":33,"command":34},"npx","npx skills add joeloverbeck\u002Fcontinuity-loom --skill improve-codebase-architecture",{"tool":36,"command":37},"bunx","bunx skills add joeloverbeck\u002Fcontinuity-loom --skill improve-codebase-architecture",{"tool":39,"command":40},"pnpm","pnpm dlx skills add joeloverbeck\u002Fcontinuity-loom --skill improve-codebase-architecture","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fcontinuity-loom\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fimprove-codebase-architecture\u002FSKILL.md",[43],{"path":44,"resourceType":45,"resourceTypeDesc":46,"sizeBytes":47,"sourceHtmlUrl":48},"HTML-REPORT.md","REFERENCE","参考文档",4423,"https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fcontinuity-loom\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fimprove-codebase-architecture\u002FHTML-REPORT.md",[50,57,62,67,74,79],{"skillNo":51,"name":52,"nameZh":6,"description":53,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":54,"categories":55,"keywords":56,"publishedAt":18},"01M758XG3F9AESCQBKYQX0GPSD","triage","Move issues and external PRs through a state machine of triage roles — categorise, verify, grill if needed, and write agent-ready briefs.","https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fcontinuity-loom\u002Fblob\u002Fmain\u002Freports\u002Fskill-evidence\u002Ftriage\u002Freviews\u002Frev_a4ce832d-b023-4fa0-8d5d-382e558791d8\u002Fcandidate\u002FSKILL.md",[],[],{"skillNo":58,"name":52,"nameZh":6,"description":53,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":59,"categories":60,"keywords":61,"publishedAt":18},"01M758XDY39PNPRY8FSXVE8SZN","https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fcontinuity-loom\u002Fblob\u002Fmain\u002Freports\u002Fskill-evidence\u002Ftriage\u002Fdecontamination\u002Fdec_2d4305f8-ee96-4376-a267-061413c4fdd6\u002Fcandidate\u002FSKILL.md",[],[],{"skillNo":63,"name":52,"nameZh":6,"description":53,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":64,"categories":65,"keywords":66,"publishedAt":18},"01M758XB4XZ6N5XP8XZXWVMVQ0","https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fcontinuity-loom\u002Fblob\u002Fmain\u002Freports\u002Fskill-evidence\u002Ftriage\u002Fdecontamination\u002Fdec_2d4305f8-ee96-4376-a267-061413c4fdd6\u002Fbaseline\u002FSKILL.md",[],[],{"skillNo":68,"name":69,"nameZh":6,"description":70,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":71,"categories":72,"keywords":73,"publishedAt":18},"01M758X7WRCGV9BM5RNEGH4MSX","to-prd","Turn the current conversation into a PRD and publish it to the project issue tracker — no requirements interview, just synthesis of what you've already discussed; the seam confirmation is the sole checkpoint.","https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fcontinuity-loom\u002Fblob\u002Fmain\u002Freports\u002Fskill-evidence\u002Fto-prd\u002Fdecontamination\u002Fdec_581ad8e4-4f9d-4aad-a6ee-59eeec05c8a4\u002Fpre-land-backup\u002FSKILL.md",[],[],{"skillNo":75,"name":69,"nameZh":6,"description":70,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":76,"categories":77,"keywords":78,"publishedAt":18},"01M758X55QJQTVHH1MWHJF2BBZ","https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fcontinuity-loom\u002Fblob\u002Fmain\u002Freports\u002Fskill-evidence\u002Fto-prd\u002Fdecontamination\u002Fdec_581ad8e4-4f9d-4aad-a6ee-59eeec05c8a4\u002Fcandidate\u002FSKILL.md",[],[],{"skillNo":80,"name":69,"nameZh":6,"description":70,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":81,"categories":82,"keywords":83,"publishedAt":18},"01M758X1Z0F2YRRHNTR3ZCY02C","https:\u002F\u002Fgithub.com\u002Fjoeloverbeck\u002Fcontinuity-loom\u002Fblob\u002Fmain\u002Freports\u002Fskill-evidence\u002Fto-prd\u002Fdecontamination\u002Fdec_581ad8e4-4f9d-4aad-a6ee-59eeec05c8a4\u002Fbaseline\u002FSKILL.md",[],[]]