[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M77K4VDAVMQ3HDK6HN72BCFN":3,"skill-same:01M77K4VDAVMQ3HDK6HN72BCFN":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},"01M77K4VDAVMQ3HDK6HN72BCFN","dress",null,"Use when adopting loom on a repo, or re-tuning an existing loom configuration.","ZPerrin\u002Floom","ZPerrin","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F6732345?v=4","01M77K4SE6DQV2GZ683DV0E7TN","01M77K4SF9C4MTN5KJM90Y895Z",0,"MIT","https:\u002F\u002Fgithub.com\u002FZPerrin\u002Floom\u002Fblob\u002Fmaster\u002Fskills\u002Fdress\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: dress\ndescription: Use when adopting loom on a repo, or re-tuning an existing loom configuration.\n---\n\n## Dress\n\nInstall or retune loom for a repo. dress decides the repo-local control surfaces: `.loom\u002Floom.toml`, optional `.loom\u002Fskills\u002F\u003Cskill>.md` overrides, managed docs, and `.loom\u002Fscripts\u002F` hooks. The plugin supplies the runtime (`doc-scan`, `doc-slicer`, `doc-linter`, `doc-stamp`, `skill-hook`); dress configures what that runtime sees.\n\n`dress` handles three repo states:\n\n- **Blank\u002Fnew:** little structure exists; propose the smallest useful loom surface.\n- **Undressed existing:** code\u002Fdocs exist but `.loom\u002Floom.toml` does not; map existing material before seeding anything.\n- **Dressed:** `.loom\u002Floom.toml` exists; read current config and overrides, then propose retuning instead of reinstalling.\n\n## Dress Control Surfaces\n\nThese are the surfaces `dress` reads or writes directly; the full key map is the [reference project](..\u002F..\u002Freferences\u002Freference-project.md), which Propose loads whole.\n\n| Surface | Dress uses it for |\n|---|---|\n| `.loom\u002Floom.toml` | read existing config; propose or write the repo control plane |\n| `.loom\u002Fskills\u002F\u003Cskill>.md` | read existing repo opinion; propose only when a skill needs repo-specific guidance |\n| `.loom\u002Fscripts\u002F*` | detect or scaffold approved deterministic hook scripts |\n| managed Markdown\u002Ffrontmatter | adopt, seed, move, exclude, or stamp docs in the managed set |\n\n## Workflow Graph\n\n```mermaid\nflowchart TD\n    survey[\"Survey - explore the repo, write nothing\"] --> propose[\"Propose - reconcile to the minimum surface\"]\n    propose --> confirm{\"Confirm - approval boundary\"}\n    confirm -->|revise| propose\n    confirm -->|approved| write[\"Write - materialize the approved set\"]\n    write --> check[\"Self-check - run checks, report drift\"]\n    check -->|\"needs an unconfirmed edit\"| propose\n```\n\n## Workflow\n\nThe gate first: a `loom gate: dress` receipt in your context says whether the repo opinion was read and which hook ran, with that opinion and that hook's output beneath it. Without one for this invocation, do the same by hand: read `.loom\u002Fskills\u002Fdress.md` if it exists and run `bash \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Fskill-hook\" dress`; exit 3 is no hook. A receipt from an earlier invocation does not count, and a hook tolerates a repeat.\n\n### 1. Survey - explore the repo, write nothing\n\nDetermine whether this repo is blank, undressed, or already dressed, then gather only the facts needed to propose the surface.\n\n- Check `.loom\u002F`: whether `.loom\u002Floom.toml` exists, which `.loom\u002Fskills\u002F\u003Cskill>.md` overrides exist, and whether hook scripts already exist under `.loom\u002Fscripts\u002F`.\n- Run `bash \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Fdoc-scan\"` to see what loom currently sees: the managed docs and the frontmatter-less candidates.\n- Run `bash \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Fdoc-slicer\"` to see the current startup slice.\n- Inspect only enough repo shape to explain the proposal: modules, build\u002Ftest signals, existing docs worth mapping, and obvious gaps.\n\n### 2. Propose - reconcile to the minimum surface\n\nReview the reference project as calibration for the minimum loom surface, then turn the survey into an adoption diff: current state -> needed loom surface -> proposed changes. Still write nothing.\n\n```bash\ncat \"${CLAUDE_PLUGIN_ROOT}\u002Freferences\u002Freference-project.md\"\n```\n\n- **Control plane.** Always propose `.loom\u002Floom.toml`; say whether it is new or a retune, and whether an existing one is valid.\n- **Managed set.** For existing Markdown, say adopt, leave unmanaged, exclude, move, or seed. `AGENTS.md` is the default agent-facing `readme`; seed only durable content, and seed `CLAUDE.md` as the one-line `@AGENTS.md` import, excluded from discovery, since Claude Code reads only the former.\n- **Startup slice.** Propose `context.slice_headers` and `context.inject_fields` from headings that exist or docs you propose to create.\n- **Overrides.** Propose `.loom\u002Fskills\u002F\u003Cskill>.md` only for repo opinion that changes a skill's DEFAULT behavior.\n- **Hooks\u002Fscripts.** Propose scripts only for deterministic behavior.\n- **Gaps.** Name what loom needs but the repo does not yet know. Do not fill unknowns with invented prose.\n\n### 3. Confirm - approval boundary\n\nDo not write until the operator approves the exact adoption diff. If the operator changes scope, revise the proposal first. Approval covers the docs, config, overrides, scripts, moves, excludes, and staged files for this pass.\n\n### 4. Write - materialize the approved set\n\nWrite one coherent pass from the approved diff.\n\n- Always write or update `.loom\u002Floom.toml`.\n- Adopt, move, seed, or exclude only the approved docs.\n- Preserve operator voice when mapping existing prose.\n- Stamp managed docs with `bash \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Fdoc-stamp\" \u003Cfile> kind=... status=... updated=\u003Ctoday>`.\n- Stage only approved files changed by dress.\n\n### 5. Self-check - run checks, report drift\n\nRun `bash \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Fdoc-linter\"`. Fix findings caused by the approved dress work. Report unrelated drift separately. If the fix requires an unapproved change, return to Propose.\n\n## Output\n\nReport what changed, what was staged, what gaps remain, and the self-check result.\n","## Dress\n\nInstall or retune loom for a repo. dress decides the repo-local control surfaces: `.loom\u002Floom.toml`, optional `.loom\u002Fskills\u002F\u003Cskill>.md` overrides, managed docs, and `.loom\u002Fscripts\u002F` hooks. The plugin supplies the runtime (`doc-scan`, `doc-slicer`, `doc-linter`, `doc-stamp`, `skill-hook`); dress configures what that runtime sees.\n\n`dress` handles three repo states:\n\n- **Blank\u002Fnew:** little structure exists; propose the smallest useful loom surface.\n- **Undressed existing:** code\u002Fdocs exist but `.loom\u002Floom.toml` does not; map existing material before seeding anything.\n- **Dressed:** `.loom\u002Floom.toml` exists; read current config and overrides, then propose retuning instead of reinstalling.\n\n## Dress Control Surfaces\n\nThese are the surfaces `dress` reads or writes directly; the full key map is the [reference project](..\u002F..\u002Freferences\u002Freference-project.md), which Propose loads whole.\n\n| Surface | Dress uses it for |\n|---|---|\n| `.loom\u002Floom.toml` | read existing config; propose or write the repo control plane |\n| `.loom\u002Fskills\u002F\u003Cskill>.md` | read existing repo opinion; propose only when a skill needs repo-specific guidance |\n| `.loom\u002Fscripts\u002F*` | detect or scaffold approved deterministic hook scripts |\n| managed Markdown\u002Ffrontmatter | adopt, seed, move, exclude, or stamp docs in the managed set |\n\n## Workflow Graph\n\n```mermaid\nflowchart TD\n    survey[\"Survey - explore the repo, write nothing\"] --> propose[\"Propose - reconcile to the minimum surface\"]\n    propose --> confirm{\"Confirm - approval boundary\"}\n    confirm -->|revise| propose\n    confirm -->|approved| write[\"Write - materialize the approved set\"]\n    write --> check[\"Self-check - run checks, report drift\"]\n    check -->|\"needs an unconfirmed edit\"| propose\n```\n\n## Workflow\n\nThe gate first: a `loom gate: dress` receipt in your context says whether the repo opinion was read and which hook ran, with that opinion and that hook's output beneath it. Without one for this invocation, do the same by hand: read `.loom\u002Fskills\u002Fdress.md` if it exists and run `bash \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Fskill-hook\" dress`; exit 3 is no hook. A receipt from an earlier invocation does not count, and a hook tolerates a repeat.\n\n### 1. Survey - explore the repo, write nothing\n\nDetermine whether this repo is blank, undressed, or already dressed, then gather only the facts needed to propose the surface.\n\n- Check `.loom\u002F`: whether `.loom\u002Floom.toml` exists, which `.loom\u002Fskills\u002F\u003Cskill>.md` overrides exist, and whether hook scripts already exist under `.loom\u002Fscripts\u002F`.\n- Run `bash \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Fdoc-scan\"` to see what loom currently sees: the managed docs and the frontmatter-less candidates.\n- Run `bash \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Fdoc-slicer\"` to see the current startup slice.\n- Inspect only enough repo shape to explain the proposal: modules, build\u002Ftest signals, existing docs worth mapping, and obvious gaps.\n\n### 2. Propose - reconcile to the minimum surface\n\nReview the reference project as calibration for the minimum loom surface, then turn the survey into an adoption diff: current state -> needed loom surface -> proposed changes. Still write nothing.\n\n```bash\ncat \"${CLAUDE_PLUGIN_ROOT}\u002Freferences\u002Freference-project.md\"\n```\n\n- **Control plane.** Always propose `.loom\u002Floom.toml`; say whether it is new or a retune, and whether an existing one is valid.\n- **Managed set.** For existing Markdown, say adopt, leave unmanaged, exclude, move, or seed. `AGENTS.md` is the default agent-facing `readme`; seed only durable content, and seed `CLAUDE.md` as the one-line `@AGENTS.md` import, excluded from discovery, since Claude Code reads only the former.\n- **Startup slice.** Propose `context.slice_headers` and `context.inject_fields` from headings that exist or docs you propose to create.\n- **Overrides.** Propose `.loom\u002Fskills\u002F\u003Cskill>.md` only for repo opinion that changes a skill's DEFAULT behavior.\n- **Hooks\u002Fscripts.** Propose scripts only for deterministic behavior.\n- **Gaps.** Name what loom needs but the repo does not yet know. Do not fill unknowns with invented prose.\n\n### 3. Confirm - approval boundary\n\nDo not write until the operator approves the exact adoption diff. If the operator changes scope, revise the proposal first. Approval covers the docs, config, overrides, scripts, moves, excludes, and staged files for this pass.\n\n### 4. Write - materialize the approved set\n\nWrite one coherent pass from the approved diff.\n\n- Always write or update `.loom\u002Floom.toml`.\n- Adopt, move, seed, or exclude only the approved docs.\n- Preserve operator voice when mapping existing prose.\n- Stamp managed docs with `bash \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Fdoc-stamp\" \u003Cfile> kind=... status=... updated=\u003Ctoday>`.\n- Stage only approved files changed by dress.\n\n### 5. Self-check - run checks, report drift\n\nRun `bash \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Fdoc-linter\"`. Fix findings caused by the approved dress work. Report unrelated drift separately. If the fix requires an unapproved change, return to Propose.\n\n## Output\n\nReport what changed, what was staged, what gaps remain, and the self-check result.\n",{"name":5,"description":7},"master","1654547be6e5893679da596a76333a780abf7a5a","skills\u002Fdress","https:\u002F\u002Fgithub.com\u002FZPerrin\u002Floom","https:\u002F\u002Fgithub.com\u002FZPerrin","2026-09-08 11:58:22",{"source":8,"skillName":5,"commands":30,"prompt":40,"zipAvailable":41},[31,34,37],{"tool":32,"command":33},"npx","npx skills add ZPerrin\u002Floom --skill dress",{"tool":35,"command":36},"bunx","bunx skills add ZPerrin\u002Floom --skill dress",{"tool":38,"command":39},"pnpm","pnpm dlx skills add ZPerrin\u002Floom --skill dress","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FZPerrin\u002Floom\u002Fblob\u002Fmaster\u002Fskills\u002Fdress\u002FSKILL.md",true,[],[44,51,58,65,72],{"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},"01M77K5BNBMX9GMZYVQ6R5RWPK","weft","Use when a loom-managed documentation set needs an editorial pass to prune, condense, reroute, and realign durable project docs with the code and the editorial ethos.","https:\u002F\u002Fgithub.com\u002FZPerrin\u002Floom\u002Fblob\u002Fmaster\u002Fskills\u002Fweft\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},"01M77K593XBM4TW4B8E9JSWTDJ","weave","Use when wrapping up a loom work session, especially when durable docs, close-out checks, merge handoff, or branch cleanup need attention.","https:\u002F\u002Fgithub.com\u002FZPerrin\u002Floom\u002Fblob\u002Fmaster\u002Fskills\u002Fweave\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},"01M77K54XGV4J50S4S99S5DQES","warp","Use when opening a unit of work in a loom managed repo, especially when orientation, branch\u002Fworktree setup, ticket context, or kickoff guidance may be needed.","https:\u002F\u002Fgithub.com\u002FZPerrin\u002Floom\u002Fblob\u002Fmaster\u002Fskills\u002Fwarp\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},"01M77K512EAS982K3E2B2FY23S","spec","Use when the user asks to spec a capability, write its requirements, or capture what the code does before changing it. Produces a repo spec of what the code does or a work spec of what one unit of work will make true, under the loom spec grammar.","https:\u002F\u002Fgithub.com\u002FZPerrin\u002Floom\u002Fblob\u002Fmaster\u002Fskills\u002Fspec\u002FSKILL.md",[],[],{"skillNo":73,"name":74,"nameZh":6,"description":75,"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},"01M77K4XPGYKFDCVBX9TVN2MBC","refine-spec","Use when the user asks whether a spec is still accurate, suspects spec drift, wants a capability's behavior captured before or after a change, or names code that has no spec. Reconciles one capability's loom spec against its tests and code and reports drift by requirement id.","https:\u002F\u002Fgithub.com\u002FZPerrin\u002Floom\u002Fblob\u002Fmaster\u002Fskills\u002Frefine-spec\u002FSKILL.md",[],[]]