[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M79A79DVZTY9AS5Q0MAG0VVN":3,"skill-same:01M79A79DVZTY9AS5Q0MAG0VVN":53},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":29,"ownerHtmlUrl":30,"ownerBlogUrl":6,"ownerDescription":31,"updatedAt":32,"install":33,"resources":46},"01M79A79DVZTY9AS5Q0MAG0VVN","cupertino-repair",null,"Use when something that already exists must be changed without breaking it -- fix, repair, adjust, extend, add to, or delete part of an existing dashboard, page, UI, CLI, library, build, kernel or platform. Trigger on 'fix this dashboard', 'this layout is broken on mobile', 'repair this', 'make X work again', 'remove this dead code', 'add a filter to this existing page', 'change this without breaking anything'. Runs objective -> baseline -> approved changeset -> scoped apply -> verify, with a hook that limits every edit to the approved item and a script-only, restorable delete. Not for designing something new: use cupertino-backwards then cupertino-council. Not for a full lifecycle pass: use cupertino-review.","Anselmoo\u002Fwerkstoff","Anselmoo","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F13209783?v=4","01M799YTB3D7RZ4P3B8CSR2X85","Anselm Hahn","01M799YTBRAM6719D9V0NRGB4R",1,"MIT","https:\u002F\u002Fgithub.com\u002FAnselmoo\u002Fwerkstoff\u002Fblob\u002Fmain\u002Fplugins\u002Fcupertino\u002Fskills\u002Fcupertino-repair\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: cupertino-repair\ndescription: \"Use when something that already exists must be changed without breaking it -- fix, repair, adjust, extend, add to, or delete part of an existing dashboard, page, UI, CLI, library, build, kernel or platform. Trigger on 'fix this dashboard', 'this layout is broken on mobile', 'repair this', 'make X work again', 'remove this dead code', 'add a filter to this existing page', 'change this without breaking anything'. Runs objective -> baseline -> approved changeset -> scoped apply -> verify, with a hook that limits every edit to the approved item and a script-only, restorable delete. Not for designing something new: use cupertino-backwards then cupertino-council. Not for a full lifecycle pass: use cupertino-review.\"\n---\n\n# cupertino-repair\n\nChange an existing system so that exactly what was asked changes, and nothing else does.\nThe other cupertino skills design something new; this one is for the opposite job, and it\nis the job they used to be misused for -- \"fix my dashboard\" routed into a redesign that\nfixed one thing and broke three.\n\nSix operations, one route, any domain: **read, verify, modify, add, create, delete**. What\ndiffers between a dashboard and a kernel is only the probe that observes it (a rendered\npage, a command, a file), never the discipline.\n\nThe PreToolUse hook enforces the route while a run is active: no source edit before an\nobjective and a baseline exist; none before the changeset is approved; each edit only to\nthe paths of the one active item; no `rm`, `mv`, `sed -i` or shell redirect into a file;\nno commit until a verdict exists; nothing under `.cupertino\u002Frepair\u002F` but the active run's\n`objective.json` (until the baseline) and `changeset.json` is yours to write; and you may not\nrun `repair_state.py approve` -- approval is the user's. A denial is the route working -- read\nits reason, do not route around it.\n\n## 1. Preflight\n\n```bash\npython3 -B \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Fpreflight_repair.py\" --target \u003Cpath the request is about>\n```\n\nStarts the run and prints what the target is (`kind`), which probes can run here, test and\nbuild commands it found, design sources (matrize tokens, CSS custom properties with their\ndark-mode blocks, chart libraries whose colours live in JS), whether the tree is clean, and a\n`verdict` of `ready` or `ready-with-gaps` with each gap named. Tell the user every gap. A web\ntarget with no browser driver can still be changed, but its checks will be reported \"not\nrun\", and a run with any \"not run\" check is never `done`.\n\n## 2. Objective -- ask, never infer\n\nWrite `.cupertino\u002Frepair\u002F\u003Crun>\u002Fobjective.json` from what the user said, then show it and ask\nthem to confirm or correct it before going further. Ask when anything is unclear; do not\ninvent acceptance criteria the user did not state.\n\n```json\n{\n  \"objective\": \"On a 390 px phone the four KPI tiles overflow sideways; make them fit, and remove legacy-chart.js which nothing uses.\",\n  \"scope\": [\"dashboard\u002F**\"],\n  \"nonGoals\": [\"no redesign of the tiles, sidebar or chart\", \"no colour changes\"],\n  \"acceptance\": [\n    {\"id\": \"A1\", \"criterion\": \"KPI row fits a 390 px viewport\",\n     \"check\": {\"kind\": \"web\", \"page\": \"dashboard\u002Findex.html\", \"width\": 390, \"scheme\": \"light\",\n               \"assert\": {\"noHorizontalOverflow\": true, \"noConsoleErrors\": true}}},\n    {\"id\": \"A2\", \"criterion\": \"legacy-chart.js is gone\",\n     \"check\": {\"kind\": \"file\", \"path\": \"dashboard\u002Flegacy-chart.js\", \"exists\": false}}\n  ],\n  \"mustPreserve\": [\n    {\"id\": \"P1\", \"criterion\": \"desktop dark mode looks the same\",\n     \"check\": {\"kind\": \"web\", \"page\": \"dashboard\u002Findex.html\", \"width\": 1280, \"scheme\": \"dark\",\n               \"assert\": {\"maxPixelDiff\": 0.5, \"noConsoleErrors\": true}}},\n    {\"id\": \"P2\", \"criterion\": \"the test suite still passes\",\n     \"check\": {\"kind\": \"command\", \"run\": \"npm test\", \"assert\": {\"exit\": 0}}}\n  ]\n}\n```\n\n`mustPreserve` is where \"do not break it\" becomes checkable. Propose at least one per thing the\nuser would be upset to lose -- the other theme, the other breakpoints, the test suite, the build,\na page that shares the stylesheet -- and let the user strike any. For a non-web target the same\nshape holds: a CLI's acceptance is a `command` probe on its output; a kernel's is a boot script\nthat greps the console (`\"run\": \"make && .\u002Fscripts\u002Fboot-qemu.sh | grep -q 'login:'\"`).\n\n```bash\npython3 -B \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Frepair_state.py\" objective\n```\n\nIt refuses a vague objective, an empty scope, or no acceptance criterion.\n\n## 3. Read -- baseline before any edit\n\n```bash\npython3 -B \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Fbaseline.py\"\n```\n\nRuns every check as things are now and hashes every file in scope. Read the target too:\nits existing design system (CSS variables, the chart library and where its palette lives,\ncomponent boundaries), its tests, and whatever the objective touches. If the user named a\ndesign source, read it now -- see [design-sources.md](references\u002Fdesign-sources.md) for a\nClaude Design System artifact, matrize tokens, or the repo's own CSS. **The existing system\nis the design system**: a repair matches what is there unless the objective says otherwise.\n\nAn acceptance check that already passes at baseline means the objective is already met or the\ncheck is wrong -- say so before planning anything.\n\n## 4. Plan -- the changeset\n\nWrite `.cupertino\u002Frepair\u002F\u003Crun>\u002Fchangeset.json`: one item per change, each the smallest change\nthat serves one acceptance criterion.\n\n```json\n{\"items\": [\n  {\"id\": \"I1\", \"op\": \"modify\", \"paths\": [\"dashboard\u002Fstyles.css\"],\n   \"rationale\": \"let the KPI row wrap; tile size untouched so desktop is identical\",\n   \"preserves\": [\"P1\"], \"checks\": [\"A1\"]},\n  {\"id\": \"I2\", \"op\": \"delete\", \"paths\": [\"dashboard\u002Flegacy-chart.js\"],\n   \"rationale\": \"superseded in March\", \"evidence\": \"no file in dashboard\u002F loads it (grep)\",\n   \"checks\": [\"A2\"]}\n]}\n```\n\n| op | means | how it is applied |\n|---|---|---|\n| `read` \u002F `verify` | inspect or run a check; changes nothing | nothing to apply |\n| `modify` | change an existing file | Write\u002FEdit, only its `paths`, file must exist |\n| `add` | extend an existing file (new function, new section) | same as modify |\n| `create` | a new file | Write, path must not exist yet |\n| `delete` | remove a file | `apply_delete.py` only -- never `rm` |\n\nPaths are concrete files inside `scope`, never globs. A `delete` carries `evidence` that nothing\ndepends on the file. Delete is a first-class repair, not something to avoid: dead code, a\nsuperseded chart, a duplicated config are part of \"fix it\" when the user's objective covers them.\n\nShow the changeset to the user as a table (id, op, paths, why) and ask them to approve it\n**themselves**. The hook denies this command when you run it, because a model that can record\nthe user's approval can approve its own plan. Give the user the line to run, in their terminal\nor in Claude Code with the `!` prefix (a `!` command runs directly, not as your tool call):\n\n```bash\n! python3 -B \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Frepair_state.py\" approve\n```\n\n(An unattended run that was started with `CUPERTINO_REPAIR_PREAPPROVED=1` in its environment -- the user's standing approval -- may run it directly.) Write the plugin path out in full when you show it -- `${CLAUDE_PLUGIN_ROOT}` is not set in the\nuser's shell. Then wait; `repair_state.py status` shows `approved` once they have run it.\nApproval is bound to the changeset's sha256: editing it afterwards voids the approval, and the\nhook will deny every source edit until the user approves again.\n\n## 5. Apply -- one item at a time\n\n```bash\npython3 -B \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Frepair_state.py\" start-item I1\n# Edit only the paths I1 names, with Write\u002FEdit.\npython3 -B \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Frepair_state.py\" finish-item I1\n\npython3 -B \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Frepair_state.py\" start-item I2\npython3 -B \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Fapply_delete.py\" --item I2\npython3 -B \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Frepair_state.py\" finish-item I2\n```\n\n`apply_delete.py` searches the scope for anything still mentioning the file and refuses if it\nfinds something; show the user those lines and pass `--accept-references` only if they agree.\nThe file moves to `trash\u002F` with its hash, so `apply_delete.py --restore` can always put it back.\n\nIf an edit turns out to need a file no item names, stop. Add an item, re-show the changeset,\napprove again. Never widen scope silently -- that is the failure this skill exists to end.\n\nWhere a change touches colour in a chart drawn on canvas or SVG by JS, keep the literal or\nresolve the variable in JS (`getComputedStyle(document.documentElement).getPropertyValue('--x')`).\n`ctx.fillStyle = \"var(--accent)\"` paints black.\n\n## 6. Verify\n\n```bash\npython3 -B \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Fverify.py\"\n```\n\nRe-runs every check, compares each with its baseline, and attributes every changed file to the\napproved item that allowed it. It writes `verdict.json` and `.cupertino\u002Frepair\u002F\u003Crun>\u002Frepair-report.html`.\n\n| verdict | meaning | next |\n|---|---|---|\n| `done` | every check passes, nothing \"not run\", every change attributed | show the report; `repair_state.py end` |\n| `partial` | an acceptance check fails or could not run | plan the gap as a new item, approve, apply, verify again |\n| `regressed` | a must-preserve check fails, a change has no item, or new console errors appeared | undo or fix the offending item first |\n\nThree verdicts without `done` HALT the run (`\"halt\": true`). Do not try a fourth time. Offer the\nuser the two honest options: restore (`apply_delete.py --restore`, then\n`repair_state.py end --abandon` and `git checkout -- \u003Cchanged paths>`), or widen the objective\ntogether. A repair that keeps retrying the same approach is guessing.\n\nAlways show the user, from the report: the verdict sentence, each check with its state and\ndetail, every change with its item, and every \"not run\" check under *what this does not prove*.\nA `done` from a run where the browser probes could not run is not available -- verify.py will\nsay `partial`.\n\n## Output\n\n```text\ncupertino-repair -- run 20261101T101500Z -- DONE (round 1)\nobjective: On a 390 px phone the four KPI tiles overflow sideways; make them fit ...\n  A1 KPI row fits a 390 px viewport        pass   scrollWidth 390 = clientWidth 390\n  A2 legacy-chart.js is gone               pass\n  P1 desktop dark mode looks the same      pass   pixel difference 0.0%\n  P2 the test suite still passes           pass   exit 0\nchanges: dashboard\u002Fstyles.css modified (I1) | dashboard\u002Flegacy-chart.js removed (I2, restorable)\nnot proven: rendering in browsers other than Chromium\nreport: .cupertino\u002Frepair\u002F20261101T101500Z\u002Frepair-report.html\n```\n","# cupertino-repair\n\nChange an existing system so that exactly what was asked changes, and nothing else does.\nThe other cupertino skills design something new; this one is for the opposite job, and it\nis the job they used to be misused for -- \"fix my dashboard\" routed into a redesign that\nfixed one thing and broke three.\n\nSix operations, one route, any domain: **read, verify, modify, add, create, delete**. What\ndiffers between a dashboard and a kernel is only the probe that observes it (a rendered\npage, a command, a file), never the discipline.\n\nThe PreToolUse hook enforces the route while a run is active: no source edit before an\nobjective and a baseline exist; none before the changeset is approved; each edit only to\nthe paths of the one active item; no `rm`, `mv`, `sed -i` or shell redirect into a file;\nno commit until a verdict exists; nothing under `.cupertino\u002Frepair\u002F` but the active run's\n`objective.json` (until the baseline) and `changeset.json` is yours to write; and you may not\nrun `repair_state.py approve` -- approval is the user's. A denial is the route working -- read\nits reason, do not route around it.\n\n## 1. Preflight\n\n```bash\npython3 -B \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Fpreflight_repair.py\" --target \u003Cpath the request is about>\n```\n\nStarts the run and prints what the target is (`kind`), which probes can run here, test and\nbuild commands it found, design sources (matrize tokens, CSS custom properties with their\ndark-mode blocks, chart libraries whose colours live in JS), whether the tree is clean, and a\n`verdict` of `ready` or `ready-with-gaps` with each gap named. Tell the user every gap. A web\ntarget with no browser driver can still be changed, but its checks will be reported \"not\nrun\", and a run with any \"not run\" check is never `done`.\n\n## 2. Objective -- ask, never infer\n\nWrite `.cupertino\u002Frepair\u002F\u003Crun>\u002Fobjective.json` from what the user said, then show it and ask\nthem to confirm or correct it before going further. Ask when anything is unclear; do not\ninvent acceptance criteria the user did not state.\n\n```json\n{\n  \"objective\": \"On a 390 px phone the four KPI tiles overflow sideways; make them fit, and remove legacy-chart.js which nothing uses.\",\n  \"scope\": [\"dashboard\u002F**\"],\n  \"nonGoals\": [\"no redesign of the tiles, sidebar or chart\", \"no colour changes\"],\n  \"acceptance\": [\n    {\"id\": \"A1\", \"criterion\": \"KPI row fits a 390 px viewport\",\n     \"check\": {\"kind\": \"web\", \"page\": \"dashboard\u002Findex.html\", \"width\": 390, \"scheme\": \"light\",\n               \"assert\": {\"noHorizontalOverflow\": true, \"noConsoleErrors\": true}}},\n    {\"id\": \"A2\", \"criterion\": \"legacy-chart.js is gone\",\n     \"check\": {\"kind\": \"file\", \"path\": \"dashboard\u002Flegacy-chart.js\", \"exists\": false}}\n  ],\n  \"mustPreserve\": [\n    {\"id\": \"P1\", \"criterion\": \"desktop dark mode looks the same\",\n     \"check\": {\"kind\": \"web\", \"page\": \"dashboard\u002Findex.html\", \"width\": 1280, \"scheme\": \"dark\",\n               \"assert\": {\"maxPixelDiff\": 0.5, \"noConsoleErrors\": true}}},\n    {\"id\": \"P2\", \"criterion\": \"the test suite still passes\",\n     \"check\": {\"kind\": \"command\", \"run\": \"npm test\", \"assert\": {\"exit\": 0}}}\n  ]\n}\n```\n\n`mustPreserve` is where \"do not break it\" becomes checkable. Propose at least one per thing the\nuser would be upset to lose -- the other theme, the other breakpoints, the test suite, the build,\na page that shares the stylesheet -- and let the user strike any. For a non-web target the same\nshape holds: a CLI's acceptance is a `command` probe on its output; a kernel's is a boot script\nthat greps the console (`\"run\": \"make && .\u002Fscripts\u002Fboot-qemu.sh | grep -q 'login:'\"`).\n\n```bash\npython3 -B \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Frepair_state.py\" objective\n```\n\nIt refuses a vague objective, an empty scope, or no acceptance criterion.\n\n## 3. Read -- baseline before any edit\n\n```bash\npython3 -B \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Fbaseline.py\"\n```\n\nRuns every check as things are now and hashes every file in scope. Read the target too:\nits existing design system (CSS variables, the chart library and where its palette lives,\ncomponent boundaries), its tests, and whatever the objective touches. If the user named a\ndesign source, read it now -- see [design-sources.md](references\u002Fdesign-sources.md) for a\nClaude Design System artifact, matrize tokens, or the repo's own CSS. **The existing system\nis the design system**: a repair matches what is there unless the objective says otherwise.\n\nAn acceptance check that already passes at baseline means the objective is already met or the\ncheck is wrong -- say so before planning anything.\n\n## 4. Plan -- the changeset\n\nWrite `.cupertino\u002Frepair\u002F\u003Crun>\u002Fchangeset.json`: one item per change, each the smallest change\nthat serves one acceptance criterion.\n\n```json\n{\"items\": [\n  {\"id\": \"I1\", \"op\": \"modify\", \"paths\": [\"dashboard\u002Fstyles.css\"],\n   \"rationale\": \"let the KPI row wrap; tile size untouched so desktop is identical\",\n   \"preserves\": [\"P1\"], \"checks\": [\"A1\"]},\n  {\"id\": \"I2\", \"op\": \"delete\", \"paths\": [\"dashboard\u002Flegacy-chart.js\"],\n   \"rationale\": \"superseded in March\", \"evidence\": \"no file in dashboard\u002F loads it (grep)\",\n   \"checks\": [\"A2\"]}\n]}\n```\n\n| op | means | how it is applied |\n|---|---|---|\n| `read` \u002F `verify` | inspect or run a check; changes nothing | nothing to apply |\n| `modify` | change an existing file | Write\u002FEdit, only its `paths`, file must exist |\n| `add` | extend an existing file (new function, new section) | same as modify |\n| `create` | a new file | Write, path must not exist yet |\n| `delete` | remove a file | `apply_delete.py` only -- never `rm` |\n\nPaths are concrete files inside `scope`, never globs. A `delete` carries `evidence` that nothing\ndepends on the file. Delete is a first-class repair, not something to avoid: dead code, a\nsuperseded chart, a duplicated config are part of \"fix it\" when the user's objective covers them.\n\nShow the changeset to the user as a table (id, op, paths, why) and ask them to approve it\n**themselves**. The hook denies this command when you run it, because a model that can record\nthe user's approval can approve its own plan. Give the user the line to run, in their terminal\nor in Claude Code with the `!` prefix (a `!` command runs directly, not as your tool call):\n\n```bash\n! python3 -B \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Frepair_state.py\" approve\n```\n\n(An unattended run that was started with `CUPERTINO_REPAIR_PREAPPROVED=1` in its environment -- the user's standing approval -- may run it directly.) Write the plugin path out in full when you show it -- `${CLAUDE_PLUGIN_ROOT}` is not set in the\nuser's shell. Then wait; `repair_state.py status` shows `approved` once they have run it.\nApproval is bound to the changeset's sha256: editing it afterwards voids the approval, and the\nhook will deny every source edit until the user approves again.\n\n## 5. Apply -- one item at a time\n\n```bash\npython3 -B \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Frepair_state.py\" start-item I1\n# Edit only the paths I1 names, with Write\u002FEdit.\npython3 -B \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Frepair_state.py\" finish-item I1\n\npython3 -B \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Frepair_state.py\" start-item I2\npython3 -B \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Fapply_delete.py\" --item I2\npython3 -B \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Frepair_state.py\" finish-item I2\n```\n\n`apply_delete.py` searches the scope for anything still mentioning the file and refuses if it\nfinds something; show the user those lines and pass `--accept-references` only if they agree.\nThe file moves to `trash\u002F` with its hash, so `apply_delete.py --restore` can always put it back.\n\nIf an edit turns out to need a file no item names, stop. Add an item, re-show the changeset,\napprove again. Never widen scope silently -- that is the failure this skill exists to end.\n\nWhere a change touches colour in a chart drawn on canvas or SVG by JS, keep the literal or\nresolve the variable in JS (`getComputedStyle(document.documentElement).getPropertyValue('--x')`).\n`ctx.fillStyle = \"var(--accent)\"` paints black.\n\n## 6. Verify\n\n```bash\npython3 -B \"${CLAUDE_PLUGIN_ROOT}\u002Fscripts\u002Fverify.py\"\n```\n\nRe-runs every check, compares each with its baseline, and attributes every changed file to the\napproved item that allowed it. It writes `verdict.json` and `.cupertino\u002Frepair\u002F\u003Crun>\u002Frepair-report.html`.\n\n| verdict | meaning | next |\n|---|---|---|\n| `done` | every check passes, nothing \"not run\", every change attributed | show the report; `repair_state.py end` |\n| `partial` | an acceptance check fails or could not run | plan the gap as a new item, approve, apply, verify again |\n| `regressed` | a must-preserve check fails, a change has no item, or new console errors appeared | undo or fix the offending item first |\n\nThree verdicts without `done` HALT the run (`\"halt\": true`). Do not try a fourth time. Offer the\nuser the two honest options: restore (`apply_delete.py --restore`, then\n`repair_state.py end --abandon` and `git checkout -- \u003Cchanged paths>`), or widen the objective\ntogether. A repair that keeps retrying the same approach is guessing.\n\nAlways show the user, from the report: the verdict sentence, each check with its state and\ndetail, every change with its item, and every \"not run\" check under *what this does not prove*.\nA `done` from a run where the browser probes could not run is not available -- verify.py will\nsay `partial`.\n\n## Output\n\n```text\ncupertino-repair -- run 20261101T101500Z -- DONE (round 1)\nobjective: On a 390 px phone the four KPI tiles overflow sideways; make them fit ...\n  A1 KPI row fits a 390 px viewport        pass   scrollWidth 390 = clientWidth 390\n  A2 legacy-chart.js is gone               pass\n  P1 desktop dark mode looks the same      pass   pixel difference 0.0%\n  P2 the test suite still passes           pass   exit 0\nchanges: dashboard\u002Fstyles.css modified (I1) | dashboard\u002Flegacy-chart.js removed (I2, restorable)\nnot proven: rendering in browsers other than Chromium\nreport: .cupertino\u002Frepair\u002F20261101T101500Z\u002Frepair-report.html\n```\n",{"name":5,"description":7},"main","6158c1e4e112c06af2c7ac01b13b2ed7214c0fcf","plugins\u002Fcupertino\u002Fskills\u002Fcupertino-repair",0,"https:\u002F\u002Fgithub.com\u002FAnselmoo\u002Fwerkstoff","https:\u002F\u002Fanselmoo.github.io\u002Fwerkstoff\u002F","https:\u002F\u002Fgithub.com\u002FAnselmoo","Likes to Develop","2026-11-02 04:20:59",{"source":8,"skillName":5,"commands":34,"prompt":44,"zipAvailable":45},[35,38,41],{"tool":36,"command":37},"npx","npx skills add Anselmoo\u002Fwerkstoff --skill cupertino-repair",{"tool":39,"command":40},"bunx","bunx skills add Anselmoo\u002Fwerkstoff --skill cupertino-repair",{"tool":42,"command":43},"pnpm","pnpm dlx skills add Anselmoo\u002Fwerkstoff --skill cupertino-repair","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FAnselmoo\u002Fwerkstoff\u002Fblob\u002Fmain\u002Fplugins\u002Fcupertino\u002Fskills\u002Fcupertino-repair\u002FSKILL.md",true,[47],{"path":48,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":51,"sourceHtmlUrl":52},"references\u002Fdesign-sources.md","REFERENCE","参考文档",4111,"https:\u002F\u002Fgithub.com\u002FAnselmoo\u002Fwerkstoff\u002Fblob\u002Fmain\u002Fplugins\u002Fcupertino\u002Fskills\u002Fcupertino-repair\u002Freferences\u002Fdesign-sources.md",[54,61,68,73,78,83],{"skillNo":55,"name":56,"nameZh":6,"description":57,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":58,"categories":59,"keywords":60,"publishedAt":19},"01M79AJTRBNBX03WWKYJ8ENM45","toy-check","A skill that checks things in the repository and reports what it finds.","https:\u002F\u002Fgithub.com\u002FAnselmoo\u002Fwerkstoff\u002Fblob\u002Fmain\u002Ftest\u002Fworkflows\u002Ffixtures\u002Ftoy-plugin\u002Fskills\u002Ftoy-check\u002FSKILL.md",[],[],{"skillNo":62,"name":63,"nameZh":6,"description":64,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":65,"categories":66,"keywords":67,"publishedAt":19},"01M79AJS292F3HTDRXWZVRSN5B","demo-check","Reports which generated files changed since the last generation run. Use when the user asks what drifted. Read-only; never edits a file.","https:\u002F\u002Fgithub.com\u002FAnselmoo\u002Fwerkstoff\u002Fblob\u002Fmain\u002Ftest\u002Fplugins\u002Ffixtures\u002Fnacharbeit-uncompleted-run\u002Fplugins\u002Fdemo\u002Fskills\u002Fdemo-check\u002FSKILL.md",[],[],{"skillNo":69,"name":63,"nameZh":6,"description":64,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":70,"categories":71,"keywords":72,"publishedAt":19},"01M79AJP4TFQXKT1AAK1WC1JYK","https:\u002F\u002Fgithub.com\u002FAnselmoo\u002Fwerkstoff\u002Fblob\u002Fmain\u002Ftest\u002Fplugins\u002Ffixtures\u002Fnacharbeit-stale-lock\u002Fplugins\u002Fdemo\u002Fskills\u002Fdemo-check\u002FSKILL.md",[],[],{"skillNo":74,"name":63,"nameZh":6,"description":64,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":75,"categories":76,"keywords":77,"publishedAt":19},"01M79AJM6Y50WP3G2BR4H0M60Z","https:\u002F\u002Fgithub.com\u002FAnselmoo\u002Fwerkstoff\u002Fblob\u002Fmain\u002Ftest\u002Fplugins\u002Ffixtures\u002Fnacharbeit-preflight-mixed\u002Fplugins\u002Fdemo\u002Fskills\u002Fdemo-check\u002FSKILL.md",[],[],{"skillNo":79,"name":63,"nameZh":6,"description":64,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":80,"categories":81,"keywords":82,"publishedAt":19},"01M79AJGR6DMT8KCCK071V78XM","https:\u002F\u002Fgithub.com\u002FAnselmoo\u002Fwerkstoff\u002Fblob\u002Fmain\u002Ftest\u002Fplugins\u002Ffixtures\u002Fnacharbeit-hooks-shape\u002Fplugins\u002Fdemo\u002Fskills\u002Fdemo-check\u002FSKILL.md",[],[],{"skillNo":84,"name":85,"nameZh":6,"description":86,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":87,"categories":88,"keywords":89,"publishedAt":19},"01M79AJBJ4CHC4PZVWKERWF3HF","zirkel-verify-assumptions","Checks exactly ONE named assumption against real evidence in at most 3 steps and reports one outcome: confidence raised (>=90), reading changed, or still unresolved. Use to resolve a single load-bearing uncertainty — typically a blocking entry surfaced by zirkel-clarify-scope: \"verify this one assumption\", \"is it actually true that X\", \"check this before we rely on it\". For multiple uncertainties, invoke once per uncertainty — the step budget is never shared.","https:\u002F\u002Fgithub.com\u002FAnselmoo\u002Fwerkstoff\u002Fblob\u002Fmain\u002Fplugins\u002Fzirkel\u002Fskills\u002Fzirkel-verify-assumptions\u002FSKILL.md",[],[]]