[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M79RH0W7F58CB1JMA34KCCXG":3,"skill-same:01M79RH0W7F58CB1JMA34KCCXG":45},{"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":6,"ownerHtmlUrl":29,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":30,"install":31,"resources":44},"01M79RH0W7F58CB1JMA34KCCXG","pr-description",null,"PR descriptions. Use whenever creating or editing a PR body; UI-affecting changes include a conditional UI Review Handoff.","Noahlw\u002Fefcc","Noahlw","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F83105194?v=4","01M79RG1S3T0FEZR3MHHY62C8G","Wong King Lok","01M79RG1SR846VYW910G2Q85Y5",0,"MIT","https:\u002F\u002Fgithub.com\u002FNoahlw\u002Fefcc\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Fpr-description\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: pr-description\ndescription: PR descriptions. Use whenever creating or editing a PR body; UI-affecting changes include a conditional UI Review Handoff.\n---\n\n# PR description template\n\nPR resolve issue(s) must auto-close on merge via GitHub closing keywords.\n\n## Rule\n\n**First line** of PR body must be closing statement, before `## Summary`:\n\n```\nCloses #\u003CN>[, #\u003CM>, ...]\n```\n\n- List every issue this PR fully resolves. Use `Closes` (not `Fixes`\u002F`Resolves` — pick one verb, keep consistent across repo).\n- Only list issues actually resolved by this PR's diff. Issue referenced but deliberately deferred (see ADR marking out of scope) goes in prose, not closing line.\n- No issues resolved (docs-only or infra-only PR) → omit line entirely, don't write `Closes` with nothing after it.\n\n## UI Review Handoff\n\nInclude this section only for a materially app-facing UI or presentation PR. Backend-only, infrastructure-only, and documentation-only PRs omit it.\n\n### Candidate\n\n- **Review level:** L1 \u002F L2 \u002F L3\n- **Baseline Story:** \u003Cdirect Storybook\u002FPlayground locator>\n- **Material Stories:** \u003Conly the useful states for this change, or `None`>\n- **Risk viewport\u002Fplatform:** \u003Conly if relevant, or `None`>\n\n### Agent self-review\n\n- [ ] Baseline inspected across W7.\n- [ ] Material states inspected at relevant risk viewports.\n- [ ] Obvious defects found during self-review were repaired.\n- [ ] No known in-scope visual defect remains.\n- [ ] Relevant focused verification is green.\n\n### Human review focus\n\n- \u003Conly genuine subjective\u002Fhigh-risk review points; use `Routine owner spot-check` for ordinary L1 work>\n\n### Visual comparison\n\n- \u003Cfor major visual changes, include useful before\u002Fafter or representative phone\u002Fdesktop\u002Frisky-state images when available; otherwise omit this subsection>\n\nLive Storybook is the primary visual review surface; images support the review. Do not duplicate the Test plan. L1 means normal AI-owned implementation plus mandatory owner Storybook spot-check; L2 is human design review triggered by design risk; L3 is actual platform\u002Fdevice\u002Fassistive-technology\u002Fmanual truth. Detailed escalation law remains in the central repository authority.\n\n## Similarity to master (LLM-judged)\n\nEvery PR body MUST include a **Similarity to master** section. It is a reviewer-facing\ncommunication aid that summarizes how much of the existing behavior\u002Foutput the PR\npreserves relative to the mainline. It is an **LLM judgment**, not a measurement —\nno scripts, no byte-diff, no pixel-diff. The writing agent reads the actual diff and\nreasons to each verdict.\n\n### Rules\n\n1. **Baseline is explicit.** Compare the PR head against `master` (resolved via\n   merge-base). Title the section with the reference, e.g.\n   `## Similarity to master (vs master @ \u003Csha>)`. Never report a number without\n   stating what it is compared against. For a stacked PR chain (`A → B → C`), compare\n   the *accumulated head* against master — reviewers see what actually lands on the\n   mainline, not just this PR's own delta. If the PR's own delta is materially\n   different from the cumulative tip, say so in one line.\n2. **Axes are agent-chosen per diff.** Do NOT use a fixed axis list. Read the diff and\n   pick the axes that are actually informative for this change. Examples of axes you\n   may choose (add or drop as the diff demands):\n   - **Visual \u002F DOM** — UI-rendering source changed? (components, styles, markup)\n   - **Report \u002F output schema** — the persisted\u002Femitted contract changed?\n   - **Report values** — the data changed? (scores, costs, computed fields)\n   - **Behavior \u002F lifecycle** — interactions, status transitions, retry\u002Fabort\u002Fcancel?\n   - **Config \u002F infra** — CI, tooling, dependencies, deployment?\n   - **Docs \u002F ADR** — documentation-only, no runtime impact?\n   Drop an axis that the diff doesn't touch rather than reporting a meaningless number.\n   This keeps the report project-agnostic: it works for any repo, not just this one.\n3. **Every verdict is an LLM judgment with a reason.** For each axis, give a similarity\n   percentage AND a one-line justification grounded in the real diff (quote the files\n   that moved). The % is an estimate, not a measurement — its authority comes from the\n   reasoning, not the number.\n4. **Honest vocabulary.** Use these labels per axis:\n   - `unchanged (100%)` — the diff does not touch this axis.\n   - `\u003CN>%` — the diff partially changes this axis; justify the estimate.\n   - `n\u002Fa` — the axis is nondeterministic output (e.g. LLM-generated prose) and a\n     similarity % would be a fake precision. Say so explicitly.\n   - `changed by design — see summary` — an intentional behavior\u002Fdata change; link it\n     to the Summary bullet, don't dress it as a similarity failure.\n   Never invent a fake % for nondeterministic content. Never blend axes into one\n   aggregate number.\n5. **Label the report as a judgment.** One line at the end: \"Similarity is an LLM\n   assessment of the diff, not a CI gate or measurement.\" This prevents anyone from\n   mistaking it for a hard verification.\n\n### Template\n\n```md\n## Similarity to master (vs master @ \u003Csha>)\n\nThis is an LLM judgment of the diff, not a measurement.\n\n- **\u003CAxis 1>**: \u003Cverdict> — \u003Creason, quoting the diff>\n- **\u003CAxis 2>**: \u003Cverdict> — \u003Creason>\n- **\u003CAxis 3>**: \u003Cverdict> — \u003Creason>\n\nSimilarity is an LLM assessment of the diff, not a CI gate or measurement.\n```\n\n## Template (full body)\n\n```md\nCloses #\u003CN>, #\u003CM>\n\n## Summary\n- \u003Cwhat changed and why, 1-3 bullets>\n\n## Similarity to master (vs master @ \u003Csha>)\n- **\u003CAxis 1>**: \u003Cverdict> — \u003Creason>\n- **\u003CAxis 2>**: \u003Cverdict> — \u003Creason>\n\nSimilarity is an LLM assessment of the diff, not a CI gate or measurement.\n\n## Test plan\n- [ ] \u003Chow this was verified>\n\n## QA Test Steps\n1. \u003Chow qa test this>\n\n🤖 Generated with [Claude Code](https:\u002F\u002Fclaude.com\u002Fclaude-code)\nhttps:\u002F\u002Fclaude.ai\u002Fcode\u002Fsession_017QWhHHbUxYJ3kusvFbqUDu\n```\n\n## Example\n\n```md\nCloses #9, #10, #15, #16\n\n## Summary\n- primaryColor\u002FwelcomeMessage\u002FavatarUrl wired into rendering per ADR-0007\n- destroy() added for SPA cleanup\n- #17 intentionally left open — deferred per ADR-0008, not part of this PR\n\n## Similarity to master (vs master @ 1a2b3c4)\n- **Visual \u002F DOM**: unchanged (100%) — no .tsx\u002F.css files in the diff.\n- **Report values**: changed by design — color tokens and welcome copy differ; see Summary.\n- **Config**: n\u002Fa — dependency\u002Fdep-only changes, no runtime surface moved.\n\nSimilarity is an LLM assessment of the diff, not a CI gate or measurement.\n\n## Test plan\n- [x] bun run typecheck passes\n- [x] verified live in headless Chromium\n\n🤖 Generated with [Claude Code](https:\u002F\u002Fclaude.com\u002Fclaude-code)\nhttps:\u002F\u002Fclaude.ai\u002Fcode\u002Fsession_017QWhHHbUxYJ3kusvFbqUDu\n```\n","# PR description template\n\nPR resolve issue(s) must auto-close on merge via GitHub closing keywords.\n\n## Rule\n\n**First line** of PR body must be closing statement, before `## Summary`:\n\n```\nCloses #\u003CN>[, #\u003CM>, ...]\n```\n\n- List every issue this PR fully resolves. Use `Closes` (not `Fixes`\u002F`Resolves` — pick one verb, keep consistent across repo).\n- Only list issues actually resolved by this PR's diff. Issue referenced but deliberately deferred (see ADR marking out of scope) goes in prose, not closing line.\n- No issues resolved (docs-only or infra-only PR) → omit line entirely, don't write `Closes` with nothing after it.\n\n## UI Review Handoff\n\nInclude this section only for a materially app-facing UI or presentation PR. Backend-only, infrastructure-only, and documentation-only PRs omit it.\n\n### Candidate\n\n- **Review level:** L1 \u002F L2 \u002F L3\n- **Baseline Story:** \u003Cdirect Storybook\u002FPlayground locator>\n- **Material Stories:** \u003Conly the useful states for this change, or `None`>\n- **Risk viewport\u002Fplatform:** \u003Conly if relevant, or `None`>\n\n### Agent self-review\n\n- [ ] Baseline inspected across W7.\n- [ ] Material states inspected at relevant risk viewports.\n- [ ] Obvious defects found during self-review were repaired.\n- [ ] No known in-scope visual defect remains.\n- [ ] Relevant focused verification is green.\n\n### Human review focus\n\n- \u003Conly genuine subjective\u002Fhigh-risk review points; use `Routine owner spot-check` for ordinary L1 work>\n\n### Visual comparison\n\n- \u003Cfor major visual changes, include useful before\u002Fafter or representative phone\u002Fdesktop\u002Frisky-state images when available; otherwise omit this subsection>\n\nLive Storybook is the primary visual review surface; images support the review. Do not duplicate the Test plan. L1 means normal AI-owned implementation plus mandatory owner Storybook spot-check; L2 is human design review triggered by design risk; L3 is actual platform\u002Fdevice\u002Fassistive-technology\u002Fmanual truth. Detailed escalation law remains in the central repository authority.\n\n## Similarity to master (LLM-judged)\n\nEvery PR body MUST include a **Similarity to master** section. It is a reviewer-facing\ncommunication aid that summarizes how much of the existing behavior\u002Foutput the PR\npreserves relative to the mainline. It is an **LLM judgment**, not a measurement —\nno scripts, no byte-diff, no pixel-diff. The writing agent reads the actual diff and\nreasons to each verdict.\n\n### Rules\n\n1. **Baseline is explicit.** Compare the PR head against `master` (resolved via\n   merge-base). Title the section with the reference, e.g.\n   `## Similarity to master (vs master @ \u003Csha>)`. Never report a number without\n   stating what it is compared against. For a stacked PR chain (`A → B → C`), compare\n   the *accumulated head* against master — reviewers see what actually lands on the\n   mainline, not just this PR's own delta. If the PR's own delta is materially\n   different from the cumulative tip, say so in one line.\n2. **Axes are agent-chosen per diff.** Do NOT use a fixed axis list. Read the diff and\n   pick the axes that are actually informative for this change. Examples of axes you\n   may choose (add or drop as the diff demands):\n   - **Visual \u002F DOM** — UI-rendering source changed? (components, styles, markup)\n   - **Report \u002F output schema** — the persisted\u002Femitted contract changed?\n   - **Report values** — the data changed? (scores, costs, computed fields)\n   - **Behavior \u002F lifecycle** — interactions, status transitions, retry\u002Fabort\u002Fcancel?\n   - **Config \u002F infra** — CI, tooling, dependencies, deployment?\n   - **Docs \u002F ADR** — documentation-only, no runtime impact?\n   Drop an axis that the diff doesn't touch rather than reporting a meaningless number.\n   This keeps the report project-agnostic: it works for any repo, not just this one.\n3. **Every verdict is an LLM judgment with a reason.** For each axis, give a similarity\n   percentage AND a one-line justification grounded in the real diff (quote the files\n   that moved). The % is an estimate, not a measurement — its authority comes from the\n   reasoning, not the number.\n4. **Honest vocabulary.** Use these labels per axis:\n   - `unchanged (100%)` — the diff does not touch this axis.\n   - `\u003CN>%` — the diff partially changes this axis; justify the estimate.\n   - `n\u002Fa` — the axis is nondeterministic output (e.g. LLM-generated prose) and a\n     similarity % would be a fake precision. Say so explicitly.\n   - `changed by design — see summary` — an intentional behavior\u002Fdata change; link it\n     to the Summary bullet, don't dress it as a similarity failure.\n   Never invent a fake % for nondeterministic content. Never blend axes into one\n   aggregate number.\n5. **Label the report as a judgment.** One line at the end: \"Similarity is an LLM\n   assessment of the diff, not a CI gate or measurement.\" This prevents anyone from\n   mistaking it for a hard verification.\n\n### Template\n\n```md\n## Similarity to master (vs master @ \u003Csha>)\n\nThis is an LLM judgment of the diff, not a measurement.\n\n- **\u003CAxis 1>**: \u003Cverdict> — \u003Creason, quoting the diff>\n- **\u003CAxis 2>**: \u003Cverdict> — \u003Creason>\n- **\u003CAxis 3>**: \u003Cverdict> — \u003Creason>\n\nSimilarity is an LLM assessment of the diff, not a CI gate or measurement.\n```\n\n## Template (full body)\n\n```md\nCloses #\u003CN>, #\u003CM>\n\n## Summary\n- \u003Cwhat changed and why, 1-3 bullets>\n\n## Similarity to master (vs master @ \u003Csha>)\n- **\u003CAxis 1>**: \u003Cverdict> — \u003Creason>\n- **\u003CAxis 2>**: \u003Cverdict> — \u003Creason>\n\nSimilarity is an LLM assessment of the diff, not a CI gate or measurement.\n\n## Test plan\n- [ ] \u003Chow this was verified>\n\n## QA Test Steps\n1. \u003Chow qa test this>\n\n🤖 Generated with [Claude Code](https:\u002F\u002Fclaude.com\u002Fclaude-code)\nhttps:\u002F\u002Fclaude.ai\u002Fcode\u002Fsession_017QWhHHbUxYJ3kusvFbqUDu\n```\n\n## Example\n\n```md\nCloses #9, #10, #15, #16\n\n## Summary\n- primaryColor\u002FwelcomeMessage\u002FavatarUrl wired into rendering per ADR-0007\n- destroy() added for SPA cleanup\n- #17 intentionally left open — deferred per ADR-0008, not part of this PR\n\n## Similarity to master (vs master @ 1a2b3c4)\n- **Visual \u002F DOM**: unchanged (100%) — no .tsx\u002F.css files in the diff.\n- **Report values**: changed by design — color tokens and welcome copy differ; see Summary.\n- **Config**: n\u002Fa — dependency\u002Fdep-only changes, no runtime surface moved.\n\nSimilarity is an LLM assessment of the diff, not a CI gate or measurement.\n\n## Test plan\n- [x] bun run typecheck passes\n- [x] verified live in headless Chromium\n\n🤖 Generated with [Claude Code](https:\u002F\u002Fclaude.com\u002Fclaude-code)\nhttps:\u002F\u002Fclaude.ai\u002Fcode\u002Fsession_017QWhHHbUxYJ3kusvFbqUDu\n```\n",{"name":5,"description":7},"main","8bcbf45239cc2a1a17807fd2d3a5e521119cf8ab","docs\u002Fskills\u002Fpr-description",1,"https:\u002F\u002Fgithub.com\u002FNoahlw\u002Fefcc","https:\u002F\u002Fgithub.com\u002FNoahlw","2026-11-09 17:42:55",{"source":8,"skillName":5,"commands":32,"prompt":42,"zipAvailable":43},[33,36,39],{"tool":34,"command":35},"npx","npx skills add Noahlw\u002Fefcc --skill pr-description",{"tool":37,"command":38},"bunx","bunx skills add Noahlw\u002Fefcc --skill pr-description",{"tool":40,"command":41},"pnpm","pnpm dlx skills add Noahlw\u002Fefcc --skill pr-description","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FNoahlw\u002Fefcc\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Fpr-description\u002FSKILL.md",true,[],[46,53,60,67,74,81],{"skillNo":47,"name":48,"nameZh":6,"description":49,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":50,"categories":51,"keywords":52,"publishedAt":19},"01M79RHRFD72JZSB0Y2ER4B7TQ","writing-plans","Use when you have a spec or requirements for a multi-step task, before touching code","https:\u002F\u002Fgithub.com\u002FNoahlw\u002Fefcc\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Fwriting-plans\u002FSKILL.md",[],[],{"skillNo":54,"name":55,"nameZh":6,"description":56,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":57,"categories":58,"keywords":59,"publishedAt":19},"01M79RHN889JQ8BY8SV6K0YG9W","writing-for-agents","Writing documents for agents. Use when creating or editing skills, or modifying AGENTS.md or CLAUDE.md.","https:\u002F\u002Fgithub.com\u002FNoahlw\u002Fefcc\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Fwriting-for-agents\u002FSKILL.md",[],[],{"skillNo":61,"name":62,"nameZh":6,"description":63,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":64,"categories":65,"keywords":66,"publishedAt":19},"01M79RHJ5K9QSMRZ1BQZMTMCA0","to-tickets","Break a plan, spec, or the current conversation into a set of tracer-bullet tickets, each declaring its blocking edges, published to the configured tracker — edges as text in one file per ticket locally, or native blocking links on a real tracker.","https:\u002F\u002Fgithub.com\u002FNoahlw\u002Fefcc\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Fto-tickets\u002FSKILL.md",[],[],{"skillNo":68,"name":69,"nameZh":6,"description":70,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":71,"categories":72,"keywords":73,"publishedAt":19},"01M79RHG1TYEX4E9BE4WWJA0C0","to-spec","Turn the current conversation into a spec and publish it to the project issue tracker — no interview, just synthesis of what you've already discussed.","https:\u002F\u002Fgithub.com\u002FNoahlw\u002Fefcc\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Fto-spec\u002FSKILL.md",[],[],{"skillNo":75,"name":76,"nameZh":6,"description":77,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":78,"categories":79,"keywords":80,"publishedAt":19},"01M79RHE5V69BD6W5JGP2QS01B","tdd","Test-driven development. Use when the user wants to build features or fix bugs test-first, mentions \"red-green-refactor\", or wants integration tests.","https:\u002F\u002Fgithub.com\u002FNoahlw\u002Fefcc\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Ftdd\u002FSKILL.md",[],[],{"skillNo":82,"name":83,"nameZh":6,"description":84,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":85,"categories":86,"keywords":87,"publishedAt":19},"01M79RGYSSWFAMZP9C37J5R59N","implement","Implement a piece of work based on a spec or set of tickets.","https:\u002F\u002Fgithub.com\u002FNoahlw\u002Fefcc\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Fimplement\u002FSKILL.md",[],[]]