[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7CZ5G5QJ7EXV9RGSRC3HGFH":3,"skill-same:01M7CZ5G5QJ7EXV9RGSRC3HGFH":60},{"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":32,"commitSha":33,"skillDirectory":34,"forksCount":35,"repositoryHtmlUrl":30,"repositoryHomepageUrl":6,"ownerHtmlUrl":36,"ownerBlogUrl":37,"ownerDescription":38,"updatedAt":39,"install":40,"resources":53},"01M7CZ5G5QJ7EXV9RGSRC3HGFH","padosoft-atomic-invariants",null,"Use this skill whenever code enforces something that must hold only once or only so many times — a single-use code or coupon, a nonce, a rate or quota limit, \"the first one wins\", a stock reservation, an idempotency key, a state transition that must happen once — and whenever the user reports that two requests both succeeded where one should have failed, a coupon was redeemed twice, a job ran twice, or a counter drifted under load. It checks that the read and the write of the invariant live in one atomic step and that a constraint backs the rule where it matters. Do not use it for general transaction design, for retry policy, or for distributed consensus.","padosoft\u002Fskills","padosoft","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F10467753?v=4","01M7CZ3FRSM851DK225E3KK2VF","01M7CZ3FSY016E21VS6EABM21V",1,"MIT","https:\u002F\u002Fgithub.com\u002Fpadosoft\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fpadosoft-atomic-invariants\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: padosoft-atomic-invariants\ndescription: >-\n  Use this skill whenever code enforces something that must hold only once or only so many times — a\n  single-use code or coupon, a nonce, a rate or quota limit, \"the first one wins\", a stock reservation, an\n  idempotency key, a state transition that must happen once — and whenever the user reports that two requests\n  both succeeded where one should have failed, a coupon was redeemed twice, a job ran twice, or a counter\n  drifted under load. It checks that the read and the write of the invariant live in one atomic step and that\n  a constraint backs the rule where it matters. Do not use it for general transaction design, for retry\n  policy, or for distributed consensus.\nlicense: MIT\ncompatibility: >-\n  Any stack with a transactional store. Examples are SQL\u002FLaravel and SQL\u002FNode; the reasoning holds wherever\n  two requests can read the same row at the same time.\nmetadata:\n  version: 0.2.0\n  author: Padosoft\n  summary: Either the invariant is recorded in the same atomic step that checked it, or it does not exist.\n  profiles: core\n  scope: global\n  repository: https:\u002F\u002Fgithub.com\u002Fpadosoft\u002Fskills\n  keywords: concurrency, transaction, race condition, idempotency, lock, unique constraint, rate limit\n---\n\n# Atomic invariants\n\n**Either the invariant is recorded in the same atomic step that checked it, or the invariant does not exist.**\n\nEverything else — a check a few lines above the write, a lock released before the update, a \"nobody clicks\ntwice\" assumption — is a race that shows up under exactly the conditions you care about: a retry, a double\ntap, a load spike, two workers on the same queue.\n\n---\n\n## 1. The shape of the bug\n\n```php\n\u002F\u002F ❌ two requests both read \"not used yet\" and both proceed\n$coupon = Coupon::whereCode($code)->firstOrFail();\nabort_if($coupon->used_at !== null, 409);\n$coupon->update(['used_at' => now()]);\n```\n\nThe gap between the read and the write is where the second request fits. It does not need to be wide: two\nrequests arriving in the same millisecond is the normal case for a double tap on a phone, a retried mobile\nrequest, or two queue workers.\n\n```php\n\u002F\u002F ✅ read and write in the same transaction, with the row locked\nDB::transaction(function () use ($code) {\n    $coupon = Coupon::whereCode($code)->lockForUpdate()->firstOrFail();\n    abort_if($coupon->used_at !== null, 409);\n    $coupon->update(['used_at' => now()]);\n});\n```\n\n**The lock is held until the invariant is recorded.** A `lockForUpdate()` whose `update()` happens after the\nclosure returns bought nothing at all.\n\n## 2. Back it with a constraint where the rule demands it\n\nA transaction is the fast path. A **unique constraint** is the one that is still true under a deploy, a\nreplica that lags, a second service that writes the same table, or a code path someone adds next year and\nforgets to wrap.\n\n```sql\nALTER TABLE coupon_redemptions ADD UNIQUE KEY uq_coupon_once (coupon_id);\n```\n\nThen treat the constraint violation as a normal outcome, not as a 500: catch it and return the same answer\nthe check would have returned. **A constraint you never expect to fire is a constraint you never handled.**\n\nThe rule of thumb: if two rows would be a *business* problem, the database says so. If they would only be\nuntidy, the transaction is enough.\n\n## 3. Check the write actually happened\n\n```php\n$affected = Coupon::whereCode($code)->whereNull('used_at')->update(['used_at' => now()]);\nif ($affected === 0) {\n    \u002F\u002F somebody else got there first — this is the answer, not an error\n}\n```\n\nA conditional update that returns its affected-row count is often simpler than a lock, and it is atomic by\nconstruction: the condition and the write are one statement. **Then read the count.** Ignoring it turns the\nwhole pattern back into a check-then-act.\n\nSame for ownership checks inside a transaction: verify, write, and confirm the row count before answering\nsuccess.\n\n## 4. Idempotency is the same rule, seen from the caller\n\nMobile networks retry. Queues retry. Users double tap. So a mutating operation that must happen once takes an\n**idempotency key generated by the caller at the moment of intent**, and the server records that key in the\nsame transaction as the effect. The second request with the same key returns the first result instead of\ndoing the work again.\n\nA key generated server-side per request is not an idempotency key — it is a request id.\n\n## 5. Where the check must not be\n\n- **Not in the client.** A disabled button, a guard in the UI, a confirmation dialog: those are user\n  experience. The invariant is enforced where the write happens.\n- **Not in a cache.** A \"have we seen this already\" lookup in a cache without a durable record is a hint, not\n  a control, and it fails open on a restart or an eviction.\n- **Not in a separate service call** whose result is used a few lines later — that is a check-then-act with a\n  network hop in the gap.\n\n## 5b. Locks and windows on a cache store\n\nA cache lock serialises work that has no row to lock, and a cache counter bounds a rate. Five details decide\nwhether they hold.\n\n- **The key is the protected resource.** Locks with different keys do not exclude each other, so a key built\n  from the request or the worker protects nothing. Build it from the thing being changed.\n- **The TTL must exceed the longest critical section.** There is no renewal: a lock that expires mid-work is\n  handed to a second holder while the first is still writing.\n- **The block-wait time is a separate, short parameter.** Waiting as long as the lock may be held pins a\n  worker for that whole time. Wait briefly, then fail or retry the job.\n- **The array store is process-local**, so it gives no exclusion across workers. A test that uses it passes\n  and proves nothing; exclusion needs a store the workers share.\n- **A rate window is `add()` then `increment()`**, never `get()` then `put()`: the second pair races, and each\n  `put()` slides the TTL so the window never closes. Two calls still leave one gap: if the key expires\n  between them, `increment()` recreates it with no TTL and the counter never resets, so a cap, once reached,\n  blocks for good. Repair it atomically: one script that increments and sets the expiry when the counter is\n  new, or, where the store has it, an expire-if-none command after the increment (Redis 7 `EXPIRE key ttl\n  NX`), which sets the expiry only when there is none and leaves the count alone. Do not repair it by\n  `put()`-ing the value you observed: it overwrites the increments made in the meantime and lets a third\n  request through a cap of two. `increment()` can also return false; compared with the cap, false reads as\n  zero and lets everything through.\n\n## 6. When failing open is the right call — and when it is not\n\nA **rate limiter** that fails open when its store is unavailable is a deliberate and usually correct choice:\nrate limiting is an availability control, and failing it closed turns a cache blip into a full outage.\nAuthorization sits upstream and does not depend on it.\n\nA **cap on cost or egress** — paid calls, outbound sends, bytes leaving the system — looks the same in code\nand is not the same control: it protects a bill or a third party, so it fails closed.\n\nA **single-use invariant** that fails open is a defect, always: it means the thing that must happen once can\nhappen twice.\n\nWrite down which of the two you are building, next to the code. The difference is not visible from the\nimplementation, and the next person will assume the wrong one.\n\n---\n\n## How to find it in a diff\n\n```bash\n# a check followed by an update, no transaction in sight\nrg -n -B2 -A6 \"(firstOrFail|first\\(\\)|find\\()\" app\u002F src\u002F | rg -n \"update\\(|save\\(\" | head\n# locks whose transaction boundary is unclear\nrg -n \"lockForUpdate|FOR UPDATE|SELECT .* FOR SHARE\" app\u002F src\u002F\n# update() whose return value is discarded\nrg -n \"^\\s*(\\\\\\$\\w+->)?update\\(\" app\u002F | rg -v \"^\\s*\\\\\\$\\w+\\s*=\"\n# single-use wording without a constraint nearby\nrg -n -i \"used_at|redeemed|consumed|one[_-]?time|nonce|single[_-]?use\" app\u002F src\u002F database\u002F\n```\n\n## Gotchas\n\n- **A transaction that spans a network call holds locks for the duration of that call.** Do the external work\n  outside; keep the invariant's read and write inside.\n- **The framework's \"unique\" validation rule is not a constraint.** It is a query before the insert, which is\n  exactly the race this skill is about.\n- **Two workers is not a hypothetical.** Any queue with concurrency above one, any horizontally scaled app,\n  any retry — the second execution is the normal case, not the edge.\n- **A test cannot usually reproduce this alone.** Test the pieces: assert the constraint exists, assert the\n  conditional update returns zero on the second call, assert the violation is handled as a business outcome.\n- **\"It has never happened\" is a statement about traffic, not about correctness.** It starts happening on the\n  day the campaign goes out.\n\n## Checklist\n\n- [ ] The read of the invariant and the write that records it are in one atomic step\n- [ ] The lock (if any) is held until after the write, inside the same transaction\n- [ ] A unique constraint backs the rule where two rows would be a business problem\n- [ ] The constraint violation is handled as an outcome, not as a 500\n- [ ] Affected-row counts are read, not discarded\n- [ ] Idempotency key generated by the caller, recorded with the effect\n- [ ] Cache locks: key is the resource, TTL above the longest section, no exclusion on the array store\n- [ ] Rate windows use `add()` then `increment()`; a counter that lost its expiry gets it back without its count\n  being rewritten (script, or expire-if-none); the return of `increment()` is checked\n- [ ] Fail-open vs fail-closed is a written decision, next to the code\n\n## Final report\n\n```\nInvariant: \u003Cwhat must hold once>\nEnforced: \u003Ctransaction + lock | conditional update + row count | unique constraint>\nConstraint present: yes \u003Cname> | no, because \u003Creason>\nViolation path: \u003Cwhat the caller gets>\nIdempotency: \u003Ckey source, where recorded> | not applicable\nFail mode under store outage: open | closed — and why\n```\n","# Atomic invariants\n\n**Either the invariant is recorded in the same atomic step that checked it, or the invariant does not exist.**\n\nEverything else — a check a few lines above the write, a lock released before the update, a \"nobody clicks\ntwice\" assumption — is a race that shows up under exactly the conditions you care about: a retry, a double\ntap, a load spike, two workers on the same queue.\n\n---\n\n## 1. The shape of the bug\n\n```php\n\u002F\u002F ❌ two requests both read \"not used yet\" and both proceed\n$coupon = Coupon::whereCode($code)->firstOrFail();\nabort_if($coupon->used_at !== null, 409);\n$coupon->update(['used_at' => now()]);\n```\n\nThe gap between the read and the write is where the second request fits. It does not need to be wide: two\nrequests arriving in the same millisecond is the normal case for a double tap on a phone, a retried mobile\nrequest, or two queue workers.\n\n```php\n\u002F\u002F ✅ read and write in the same transaction, with the row locked\nDB::transaction(function () use ($code) {\n    $coupon = Coupon::whereCode($code)->lockForUpdate()->firstOrFail();\n    abort_if($coupon->used_at !== null, 409);\n    $coupon->update(['used_at' => now()]);\n});\n```\n\n**The lock is held until the invariant is recorded.** A `lockForUpdate()` whose `update()` happens after the\nclosure returns bought nothing at all.\n\n## 2. Back it with a constraint where the rule demands it\n\nA transaction is the fast path. A **unique constraint** is the one that is still true under a deploy, a\nreplica that lags, a second service that writes the same table, or a code path someone adds next year and\nforgets to wrap.\n\n```sql\nALTER TABLE coupon_redemptions ADD UNIQUE KEY uq_coupon_once (coupon_id);\n```\n\nThen treat the constraint violation as a normal outcome, not as a 500: catch it and return the same answer\nthe check would have returned. **A constraint you never expect to fire is a constraint you never handled.**\n\nThe rule of thumb: if two rows would be a *business* problem, the database says so. If they would only be\nuntidy, the transaction is enough.\n\n## 3. Check the write actually happened\n\n```php\n$affected = Coupon::whereCode($code)->whereNull('used_at')->update(['used_at' => now()]);\nif ($affected === 0) {\n    \u002F\u002F somebody else got there first — this is the answer, not an error\n}\n```\n\nA conditional update that returns its affected-row count is often simpler than a lock, and it is atomic by\nconstruction: the condition and the write are one statement. **Then read the count.** Ignoring it turns the\nwhole pattern back into a check-then-act.\n\nSame for ownership checks inside a transaction: verify, write, and confirm the row count before answering\nsuccess.\n\n## 4. Idempotency is the same rule, seen from the caller\n\nMobile networks retry. Queues retry. Users double tap. So a mutating operation that must happen once takes an\n**idempotency key generated by the caller at the moment of intent**, and the server records that key in the\nsame transaction as the effect. The second request with the same key returns the first result instead of\ndoing the work again.\n\nA key generated server-side per request is not an idempotency key — it is a request id.\n\n## 5. Where the check must not be\n\n- **Not in the client.** A disabled button, a guard in the UI, a confirmation dialog: those are user\n  experience. The invariant is enforced where the write happens.\n- **Not in a cache.** A \"have we seen this already\" lookup in a cache without a durable record is a hint, not\n  a control, and it fails open on a restart or an eviction.\n- **Not in a separate service call** whose result is used a few lines later — that is a check-then-act with a\n  network hop in the gap.\n\n## 5b. Locks and windows on a cache store\n\nA cache lock serialises work that has no row to lock, and a cache counter bounds a rate. Five details decide\nwhether they hold.\n\n- **The key is the protected resource.** Locks with different keys do not exclude each other, so a key built\n  from the request or the worker protects nothing. Build it from the thing being changed.\n- **The TTL must exceed the longest critical section.** There is no renewal: a lock that expires mid-work is\n  handed to a second holder while the first is still writing.\n- **The block-wait time is a separate, short parameter.** Waiting as long as the lock may be held pins a\n  worker for that whole time. Wait briefly, then fail or retry the job.\n- **The array store is process-local**, so it gives no exclusion across workers. A test that uses it passes\n  and proves nothing; exclusion needs a store the workers share.\n- **A rate window is `add()` then `increment()`**, never `get()` then `put()`: the second pair races, and each\n  `put()` slides the TTL so the window never closes. Two calls still leave one gap: if the key expires\n  between them, `increment()` recreates it with no TTL and the counter never resets, so a cap, once reached,\n  blocks for good. Repair it atomically: one script that increments and sets the expiry when the counter is\n  new, or, where the store has it, an expire-if-none command after the increment (Redis 7 `EXPIRE key ttl\n  NX`), which sets the expiry only when there is none and leaves the count alone. Do not repair it by\n  `put()`-ing the value you observed: it overwrites the increments made in the meantime and lets a third\n  request through a cap of two. `increment()` can also return false; compared with the cap, false reads as\n  zero and lets everything through.\n\n## 6. When failing open is the right call — and when it is not\n\nA **rate limiter** that fails open when its store is unavailable is a deliberate and usually correct choice:\nrate limiting is an availability control, and failing it closed turns a cache blip into a full outage.\nAuthorization sits upstream and does not depend on it.\n\nA **cap on cost or egress** — paid calls, outbound sends, bytes leaving the system — looks the same in code\nand is not the same control: it protects a bill or a third party, so it fails closed.\n\nA **single-use invariant** that fails open is a defect, always: it means the thing that must happen once can\nhappen twice.\n\nWrite down which of the two you are building, next to the code. The difference is not visible from the\nimplementation, and the next person will assume the wrong one.\n\n---\n\n## How to find it in a diff\n\n```bash\n# a check followed by an update, no transaction in sight\nrg -n -B2 -A6 \"(firstOrFail|first\\(\\)|find\\()\" app\u002F src\u002F | rg -n \"update\\(|save\\(\" | head\n# locks whose transaction boundary is unclear\nrg -n \"lockForUpdate|FOR UPDATE|SELECT .* FOR SHARE\" app\u002F src\u002F\n# update() whose return value is discarded\nrg -n \"^\\s*(\\\\\\$\\w+->)?update\\(\" app\u002F | rg -v \"^\\s*\\\\\\$\\w+\\s*=\"\n# single-use wording without a constraint nearby\nrg -n -i \"used_at|redeemed|consumed|one[_-]?time|nonce|single[_-]?use\" app\u002F src\u002F database\u002F\n```\n\n## Gotchas\n\n- **A transaction that spans a network call holds locks for the duration of that call.** Do the external work\n  outside; keep the invariant's read and write inside.\n- **The framework's \"unique\" validation rule is not a constraint.** It is a query before the insert, which is\n  exactly the race this skill is about.\n- **Two workers is not a hypothetical.** Any queue with concurrency above one, any horizontally scaled app,\n  any retry — the second execution is the normal case, not the edge.\n- **A test cannot usually reproduce this alone.** Test the pieces: assert the constraint exists, assert the\n  conditional update returns zero on the second call, assert the violation is handled as a business outcome.\n- **\"It has never happened\" is a statement about traffic, not about correctness.** It starts happening on the\n  day the campaign goes out.\n\n## Checklist\n\n- [ ] The read of the invariant and the write that records it are in one atomic step\n- [ ] The lock (if any) is held until after the write, inside the same transaction\n- [ ] A unique constraint backs the rule where two rows would be a business problem\n- [ ] The constraint violation is handled as an outcome, not as a 500\n- [ ] Affected-row counts are read, not discarded\n- [ ] Idempotency key generated by the caller, recorded with the effect\n- [ ] Cache locks: key is the resource, TTL above the longest section, no exclusion on the array store\n- [ ] Rate windows use `add()` then `increment()`; a counter that lost its expiry gets it back without its count\n  being rewritten (script, or expire-if-none); the return of `increment()` is checked\n- [ ] Fail-open vs fail-closed is a written decision, next to the code\n\n## Final report\n\n```\nInvariant: \u003Cwhat must hold once>\nEnforced: \u003Ctransaction + lock | conditional update + row count | unique constraint>\nConstraint present: yes \u003Cname> | no, because \u003Creason>\nViolation path: \u003Cwhat the caller gets>\nIdempotency: \u003Ckey source, where recorded> | not applicable\nFail mode under store outage: open | closed — and why\n```\n",{"name":5,"license":14,"metadata":23,"description":7,"compatibility":31},{"scope":24,"author":25,"summary":26,"version":27,"keywords":28,"profiles":29,"repository":30},"global","Padosoft","Either the invariant is recorded in the same atomic step that checked it, or it does not exist.","0.2.0","concurrency, transaction, race condition, idempotency, lock, unique constraint, rate limit","core","https:\u002F\u002Fgithub.com\u002Fpadosoft\u002Fskills","Any stack with a transactional store. Examples are SQL\u002FLaravel and SQL\u002FNode; the reasoning holds wherever two requests can read the same row at the same time.","main","8ff266762b08cab7f5be5551fa221309dc9db48b","skills\u002Fpadosoft-atomic-invariants",0,"https:\u002F\u002Fgithub.com\u002Fpadosoft","https:\u002F\u002Fwww.padosoft.com","Software House","2026-10-19 01:02:17",{"source":8,"skillName":5,"commands":41,"prompt":51,"zipAvailable":52},[42,45,48],{"tool":43,"command":44},"npx","npx skills add padosoft\u002Fskills --skill padosoft-atomic-invariants",{"tool":46,"command":47},"bunx","bunx skills add padosoft\u002Fskills --skill padosoft-atomic-invariants",{"tool":49,"command":50},"pnpm","pnpm dlx skills add padosoft\u002Fskills --skill padosoft-atomic-invariants","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fpadosoft\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fpadosoft-atomic-invariants\u002FSKILL.md",true,[54],{"path":55,"resourceType":56,"resourceTypeDesc":57,"sizeBytes":58,"sourceHtmlUrl":59},"evals\u002Fqueries.json","REFERENCE","参考文档",1649,"https:\u002F\u002Fgithub.com\u002Fpadosoft\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fpadosoft-atomic-invariants\u002Fevals\u002Fqueries.json",[61,68,74,81,88,94],{"skillNo":62,"name":63,"nameZh":6,"description":64,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":65,"categories":66,"keywords":67,"publishedAt":18},"01M7CZNKCTMWK7F2J6KKPAR8YM","report-opus","Writes the weekly status report from the team's tickets. Use when the user asks for the weekly report or a status summary. Do not use it for incident reports.","https:\u002F\u002Fgithub.com\u002Fpadosoft\u002Fskills\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskill-audit\u002Fproject\u002Fdot-claude\u002Fskills\u002Freport-opus\u002FSKILL.md",[],[],{"skillNo":69,"name":70,"nameZh":6,"description":64,"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},"01M7CZNFEN8XK6S8NC6KGREYGS","report-haiku","https:\u002F\u002Fgithub.com\u002Fpadosoft\u002Fskills\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskill-audit\u002Fproject\u002Fdot-claude\u002Fskills\u002Freport-haiku\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},"01M7CZND182TKGGM7R4134ND00","kitt-stub","Builds the guided-tour reference for an admin page. Use when the user asks for the tour reference. Trigger \"tour\", \"reference\", \"guided tour\".","https:\u002F\u002Fgithub.com\u002Fpadosoft\u002Fskills\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskill-audit\u002Fproject\u002Fdot-claude\u002Fskills\u002Fkitt-stub\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},"01M7CZN9F2A6BFTFW5RSRNAGB1","good-skill","Generates the release notes of a project from its merged pull requests. Use this skill when the user asks for release notes, a changelog entry or a summary of what shipped since the last tag. Do not use it to write commit messages, that is the job of a commit skill.","https:\u002F\u002Fgithub.com\u002Fpadosoft\u002Fskills\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskill-audit\u002Fproject\u002Fdot-claude\u002Fskills\u002Fgood-skill\u002FSKILL.md",[],[],{"skillNo":89,"name":90,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":91,"categories":92,"keywords":93,"publishedAt":18},"01M7CZN6BJASPD4DH8DFR63GQF","broken-frontmatter","https:\u002F\u002Fgithub.com\u002Fpadosoft\u002Fskills\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskill-audit\u002Fproject\u002Fdot-claude\u002Fskills\u002Fbroken-frontmatter\u002FSKILL.md",[],[],{"skillNo":95,"name":96,"nameZh":6,"description":97,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":98,"categories":99,"keywords":100,"publishedAt":18},"01M7CZN498TRNC30GSYMMAZNXM","Bad_Skill","I can help you with reports.","https:\u002F\u002Fgithub.com\u002Fpadosoft\u002Fskills\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskill-audit\u002Fproject\u002Fdot-claude\u002Fskills\u002Fbad-skill\u002FSKILL.md",[],[]]