[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7DF020ZPECMZWBRVD6REYHJ":3,"skill-same:01M7DF020ZPECMZWBRVD6REYHJ":51},{"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},"01M7DF020ZPECMZWBRVD6REYHJ","backend-performance",null,"Measure and improve latency, throughput or resource use in server-side request handling, background processing or data access. Use when profiling or a server-side budget identifies a backend bottleneck.","andersonmrodrigues\u002Fpaved","andersonmrodrigues","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F13907380?v=4","01M7DEV3FP2VMQZ3FP2FP1TK1C","Anderson Rodrigues","01M7DEV3GC2QAQBW4FMNT98H48",1,"MIT","https:\u002F\u002Fgithub.com\u002Fandersonmrodrigues\u002Fpaved\u002Fblob\u002Fmain\u002Fplugins\u002Fclaude\u002Fpaved\u002Fruntime\u002Fpaved-core\u002Fcore\u002Fskills\u002Fperformance\u002Fbackend-performance\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: backend-performance\ndescription: >-\n  Measure and improve latency, throughput or resource use in server-side request handling, background processing or data access. Use when profiling or a server-side budget identifies a backend bottleneck.\n---\n\n# Backend performance\n\n## When to use\n\nIn the `planning` and `implementation` phases of a `performance` run, once `profiling` has located\nthe bottleneck in server-side code. It always starts from a measurement.\n\n## Required context\n\nThe baseline measurement and profile from `profiling`. Useful: the architecture context\n(data stores, caches, remote services on the path) and the integrations context.\n\n## Preconditions\n\n- A baseline measurement and a profile exist.\n- The metric is a percentile, not only an average: tail latency is what users feel under\n  load.\n\n## Procedure\n\n1. **Classify the time** the profile shows: computing, waiting on storage, waiting on\n   remote calls, waiting on locks or pools. Each class has different fixes.\n2. **Look for the usual causes**, in this order, because they are common and cheap to\n   confirm:\n   - one query or remote call per item inside a loop, instead of one for the batch;\n   - data loaded and then discarded: unused columns, unused relations, missing limits;\n   - lookups the data store cannot serve from an index;\n   - remote calls made one after another that do not depend on each other;\n   - pools (connections, threads) too small for the load, so requests queue;\n   - repeated work on unchanged input that a cache could serve.\n3. **Fix one cause**, then measure again with `profiling`.\n4. **For caching**, decide before writing it: what the key is, how entries are\n   invalidated, how stale a result may be, what happens on a miss storm. A cache\n   without an invalidation answer is a correctness bug waiting.\n5. **For concurrency**, check shared state, ordering assumptions and limits on the\n   collaborators you now call in parallel.\n6. **Check the effect on collaborators**: a faster service can overload the next one.\n\n## Tools\n\nNone. Measurement tools come from `profiling`.\n\n## Rules\n\nNone beyond those that apply by scope; the measurement rule is enforced through\n`profiling`.\n\n## Verification\n\nRequired: `performance`. Recommended: `integration`, to show data access and remote\ncalls still behave the same.\n\n## Evidence\n\nBaseline and result measurements for the changed path.\n\n## Completion criteria\n\n- The measured percentile meets the target, measured as the baseline was.\n- Cache invalidation and concurrency limits are stated for any cache or parallelism\n  added.\n\n## Failure modes\n\n| Failure | Signal | Response |\n|---|---|---|\n| Stale data | Cache serves results that should have changed | Fix invalidation before keeping the cache |\n| Load moved, not removed | Downstream latency or errors rise | Measure the collaborator; add limits |\n| Average improved, tail did not | Median drops, high percentile unchanged | Profile the slow requests specifically |\n| Contention introduced | Throughput drops under parallel load | Revisit shared state and pool sizes |\n\n## References\n\nNone.\n","# Backend performance\n\n## When to use\n\nIn the `planning` and `implementation` phases of a `performance` run, once `profiling` has located\nthe bottleneck in server-side code. It always starts from a measurement.\n\n## Required context\n\nThe baseline measurement and profile from `profiling`. Useful: the architecture context\n(data stores, caches, remote services on the path) and the integrations context.\n\n## Preconditions\n\n- A baseline measurement and a profile exist.\n- The metric is a percentile, not only an average: tail latency is what users feel under\n  load.\n\n## Procedure\n\n1. **Classify the time** the profile shows: computing, waiting on storage, waiting on\n   remote calls, waiting on locks or pools. Each class has different fixes.\n2. **Look for the usual causes**, in this order, because they are common and cheap to\n   confirm:\n   - one query or remote call per item inside a loop, instead of one for the batch;\n   - data loaded and then discarded: unused columns, unused relations, missing limits;\n   - lookups the data store cannot serve from an index;\n   - remote calls made one after another that do not depend on each other;\n   - pools (connections, threads) too small for the load, so requests queue;\n   - repeated work on unchanged input that a cache could serve.\n3. **Fix one cause**, then measure again with `profiling`.\n4. **For caching**, decide before writing it: what the key is, how entries are\n   invalidated, how stale a result may be, what happens on a miss storm. A cache\n   without an invalidation answer is a correctness bug waiting.\n5. **For concurrency**, check shared state, ordering assumptions and limits on the\n   collaborators you now call in parallel.\n6. **Check the effect on collaborators**: a faster service can overload the next one.\n\n## Tools\n\nNone. Measurement tools come from `profiling`.\n\n## Rules\n\nNone beyond those that apply by scope; the measurement rule is enforced through\n`profiling`.\n\n## Verification\n\nRequired: `performance`. Recommended: `integration`, to show data access and remote\ncalls still behave the same.\n\n## Evidence\n\nBaseline and result measurements for the changed path.\n\n## Completion criteria\n\n- The measured percentile meets the target, measured as the baseline was.\n- Cache invalidation and concurrency limits are stated for any cache or parallelism\n  added.\n\n## Failure modes\n\n| Failure | Signal | Response |\n|---|---|---|\n| Stale data | Cache serves results that should have changed | Fix invalidation before keeping the cache |\n| Load moved, not removed | Downstream latency or errors rise | Measure the collaborator; add limits |\n| Average improved, tail did not | Median drops, high percentile unchanged | Profile the slow requests specifically |\n| Contention introduced | Throughput drops under parallel load | Revisit shared state and pool sizes |\n\n## References\n\nNone.\n",{"name":5,"description":7},"main","1cd80f05aff924ea609012e864fa5c22c16c7d2d","plugins\u002Fclaude\u002Fpaved\u002Fruntime\u002Fpaved-core\u002Fcore\u002Fskills\u002Fperformance\u002Fbackend-performance",0,"https:\u002F\u002Fgithub.com\u002Fandersonmrodrigues\u002Fpaved","https:\u002F\u002Fgithub.com\u002Fandersonmrodrigues","2026-11-14 20:09:28",{"source":8,"skillName":5,"commands":32,"prompt":42,"zipAvailable":43},[33,36,39],{"tool":34,"command":35},"npx","npx skills add andersonmrodrigues\u002Fpaved --skill backend-performance",{"tool":37,"command":38},"bunx","bunx skills add andersonmrodrigues\u002Fpaved --skill backend-performance",{"tool":40,"command":41},"pnpm","pnpm dlx skills add andersonmrodrigues\u002Fpaved --skill backend-performance","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fandersonmrodrigues\u002Fpaved\u002Fblob\u002Fmain\u002Fplugins\u002Fclaude\u002Fpaved\u002Fruntime\u002Fpaved-core\u002Fcore\u002Fskills\u002Fperformance\u002Fbackend-performance\u002FSKILL.md",true,[45],{"path":46,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":49,"sourceHtmlUrl":50},"skill.yaml","REFERENCE","参考文档",355,"https:\u002F\u002Fgithub.com\u002Fandersonmrodrigues\u002Fpaved\u002Fblob\u002Fmain\u002Fplugins\u002Fclaude\u002Fpaved\u002Fruntime\u002Fpaved-core\u002Fcore\u002Fskills\u002Fperformance\u002Fbackend-performance\u002Fskill.yaml",[52,59,65,71,77,83],{"skillNo":53,"name":54,"nameZh":6,"description":55,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":56,"categories":57,"keywords":58,"publishedAt":19},"01M7DFBWJT4T1MPBHTWT19PBEK","sample-skill","Does a sample task. Use when a sample is needed.","https:\u002F\u002Fgithub.com\u002Fandersonmrodrigues\u002Fpaved\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskills\u002Fvalid\u002Fsample-skill\u002FSKILL.md",[],[],{"skillNo":60,"name":61,"nameZh":6,"description":55,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":62,"categories":63,"keywords":64,"publishedAt":19},"01M7DFBS6WTK30PBSW4ZS0NPWT","unexplained-contract","https:\u002F\u002Fgithub.com\u002Fandersonmrodrigues\u002Fpaved\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskills\u002Finvalid\u002Funexplained-contract\u002FSKILL.md",[],[],{"skillNo":66,"name":67,"nameZh":6,"description":55,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":68,"categories":69,"keywords":70,"publishedAt":19},"01M7DFBPJGG170V9WTMFWY73R4","shell-block","https:\u002F\u002Fgithub.com\u002Fandersonmrodrigues\u002Fpaved\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskills\u002Finvalid\u002Fshell-block\u002FSKILL.md",[],[],{"skillNo":72,"name":73,"nameZh":6,"description":55,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":74,"categories":75,"keywords":76,"publishedAt":19},"01M7DFBMZYR20JPTXNE5A8EFG6","project-identifier","https:\u002F\u002Fgithub.com\u002Fandersonmrodrigues\u002Fpaved\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskills\u002Finvalid\u002Fproject-identifier\u002FSKILL.md",[],[],{"skillNo":78,"name":79,"nameZh":6,"description":55,"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},"01M7DFBJ8J533DAKA3FTF54SX3","oversized","https:\u002F\u002Fgithub.com\u002Fandersonmrodrigues\u002Fpaved\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskills\u002Finvalid\u002Foversized\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},"01M7DFBFAD0RYCRHTZ2JW4RX7E","no-activation","Does a sample task.","https:\u002F\u002Fgithub.com\u002Fandersonmrodrigues\u002Fpaved\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fskills\u002Finvalid\u002Fno-activation\u002FSKILL.md",[],[]]