[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M698QP1E1CYMK1YGA7R690RN":3,"skill-same:01M698QP1E1CYMK1YGA7R690RN":66},{"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":22,"publishedAt":38,"linkOnly":39,"markdownContent":40,"markdownBody":41,"frontMatter":42,"refName":43,"commitSha":44,"skillDirectory":45,"forksCount":46,"repositoryHtmlUrl":47,"repositoryHomepageUrl":48,"ownerHtmlUrl":49,"ownerBlogUrl":48,"ownerDescription":50,"updatedAt":51,"install":52,"resources":65},"01M698QP1E1CYMK1YGA7R690RN","test-audit",null,"Write or review Windows regression tests for observable contracts, independent expectations, faithful fixtures, and distinct coverage without weakening native proof or architecture guards.","openclaw\u002Fopenclaw-windows-node","openclaw","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F252820863?v=4","01M698P3FSAZ6B4R928ZF19KPT","01M698Q6SND6RHZDPP8ZM0NSP7",2198,"MIT","https:\u002F\u002Fgithub.com\u002Fopenclaw\u002Fopenclaw-windows-node\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Ftest-audit\u002FSKILL.md",[17],{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000008","测试与质量","SCENARIO","使用场景",[23,24,25,26,27,28,29,30,31,32,33,34,35,36,37],"test audit","test","audit","coverage","windows","for","regression","tests","architecture","review","fixtures","native","write","proof","contracts","2026-11-10 10:12:30",false,"---\nname: test-audit\ndescription: Write or review Windows regression tests for observable contracts, independent expectations, faithful fixtures, and distinct coverage without weakening native proof or architecture guards.\n---\n\n# Windows test audit\n\nUse when adding, changing, or reviewing tests. A review does not authorize\ntest deletion or production refactoring.\n\n## Authoring gate\n\nAnswer these questions:\n\n1. What observable behavior or independent contract does it protect?\n2. What credible regression makes it fail, and does it fail for that reason\n   rather than a broken fixture?\n3. Which production owner is exercised? Why does existing coverage miss this\n   risk?\n4. Are the expected values independent of the implementation under test?\n5. Does another layer add a distinct transport, persistence, lifecycle,\n   permission, or rendering risk, rather than repeat the same assertion?\n6. Is this the cheapest faithful boundary, or does it add a process, desktop,\n   WSL, installer, Gateway, or native capability boot that a smaller owner test\n   can cover?\n\nPrefer an existing table, fixture, or owner regression over another process\nboot. Avoid mocks supplying the behavior being asserted, production helpers\ncomputing their own expected values, and test-only exports without a genuine\nproduction boundary.\n\nRead [regression authoring](..\u002F..\u002F..\u002Fdocs\u002FTEST_COVERAGE.md#writing-a-regression-test)\nand use the existing [fixture guide](..\u002F..\u002F..\u002Fdocs\u002FGATEWAY_FIXTURE_TESTING.md)\nwhen the behavior crosses the application boundary.\n\n## Choose the production owner\n\nKeep one primary regression at the strongest faithful owner boundary:\n\n| Boundary | Primary project or lane |\n| --- | --- |\n| Gateway registry, credentials, connection lifecycle, migration admission\u002Frecords and persisted Local AI runtime state | `OpenClaw.Connection.Tests` |\n| Protocol, MCP, capabilities, command policy and shared models | `OpenClaw.Shared.Tests` |\n| Tray policy, settings, onboarding helpers, installer source contracts and update checks\u002Fbackends | `OpenClaw.Tray.Tests` |\n| Setup engine, WSL installation and Local AI setup\u002Frecovery policy | `OpenClaw.SetupEngine.Tests` |\n| Shipped `winnode` arguments, JSON output and command documentation | `OpenClaw.WinNode.Cli.Tests` |\n| Functional UI element graphs, factories and hook state without mounted native controls | `OpenClawTray.FunctionalUI.Tests` |\n| Mounted native WinUI controls, UI Automation, accessibility and focused input | `OpenClaw.Tray.UITests` |\n| Clean install, upgrade, uninstall and Store migration runtime behavior | The applicable formal validation path or proof pool |\n| Real process, Gateway-mediated setup\u002Fconnect or MXC behavior | Integration, E2E or named proof lane |\n\nAdd a second layer only when it proves a different failure mode. Local MCP\noutput does not prove Gateway pairing. A state snapshot is not a rendered UI\nassertion. Synthetic Gateway proof is not live provider, MXC, installer, Store,\nor production credential-resolution proof.\n\n## Reject low-value proof\n\nReject or rewrite tests that rely on:\n\n- self-comparisons, copied inventories, subject-generated expected values, or\n  the same helper on both sides of an assertion;\n- implementation names, private call shape, or exact delegation counts unless\n  that ownership\u002Fdelegation boundary is itself the protected contract;\n- a mock or fake that supplies the result, ordering, persistence receipt, or\n  policy decision being asserted;\n- another test at every crossed layer without a distinct transport,\n  persistence, permission, lifecycle, packaging, or rendering risk;\n- production flags or hooks that bypass owner logic, or exports\u002Fvisibility\n  added only to expose internal behavior without a genuine production boundary;\n- a broad source grep that pins incidental identifiers instead of a concrete\n  prohibited pattern or independent shipped inventory.\n\nNegative controls must fail at the intended guard, not during fixture setup.\nTest names must describe the behavior actually exercised. If a production seam\nexists only for obsolete tests and has no production-boundary or architecture\nrole, remove it with the tests when the task authorizes production changes;\notherwise report it rather than expanding it. Do not reject a narrow adapter\nover process, WSL, OS, network, clock, or filesystem boundaries merely because\ntests supply its only alternate implementation. Existing test assembly\nvisibility and ledger-recorded seams remain governed by their real contracts\nand retirement conditions.\n\n## Timing and determinism checklist\n\nFor native UI, loopback, process, and fixture tests, check each item against the\n[timing rules](..\u002F..\u002F..\u002Fdocs\u002FTEST_COVERAGE.md#deterministic-native-and-timing-sensitive-tests):\n\n- Readiness waits on an observable condition, not a delay, sleep, or fixed\n  settle count.\n- UI Automation elements are acquired inside the wait and used from that\n  capture, not looked up once with `!` or a throwing helper.\n- Geometry is asserted only after bounds stop moving, with an explicit\n  tolerance.\n- A polling predicate keeps polling while ownership is unproven and fails\n  fast only on positive foreign evidence.\n- Fixture data sets every value the assertion depends on; ports and other\n  inputs are deterministic.\n- Hang guards are named, environment-overridable constants with headroom for\n  a loaded runner. Latency-bounded work uses a dedicated thread, not the shared\n  thread pool.\n- A surviving flake is investigated as a possible product race before the\n  wait is changed.\n\nThe `TestTimingPatternGuard` ratchet rejects new literal short hang guards,\nfixed delays, and one-shot UI Automation lookups. Do not raise its baseline to\nadmit new code; use the named helper or a justified\n`\u002F\u002F test-timing-exception: \u003Creason>` comment. A rename or move may carry an\nexisting baseline entry to the new file when the rule totals stay unchanged.\nWhen a change removes occurrences, lower the baseline in the same change with\n`.\\scripts\\validate-test-timing-patterns.ps1 -LowerBaseline`.\n\n## Preserve valuable contracts\n\n- Keep independent protocol, CLI, storage, migration, defaults, packaging,\n  update, security, architecture, generated-resource, and user-visible\n  ordering contracts.\n- Static does not mean disposable. Source-shape guards are appropriate for a\n  concrete prohibited pattern with the\n  [ledger's retirement condition](..\u002F..\u002F..\u002Fdocs\u002FARCHITECTURE.md#how-to-use-this-document).\n  Installer source assertions remain distinct from clean install, upgrade, and\n  uninstall proof. The shipped `winnode` command-to-registry contract must\n  remain covered.\n- Use isolated settings, loopback listeners, owned processes, and existing\n  `TempDirectory`, `EnvironmentScope`, `CliHarness`, and Gateway fixture\n  helpers. Fakes may provide external input or record calls, but must not\n  implement the behavior under test.\n- Do not introduce arbitrary sleeps, blind polling, redundant process boots,\n  or parallel native input tests on a shared desktop. Reuse suite-level\n  fixtures when they preserve the same contract and isolation. Use a fresh\n  process when startup, process-global state, ownership\u002Ftermination, or an\n  independent profile is the boundary under test.\n- Do not make a failure green through weaker assertions, hidden skips,\n  blind retries, broadened mocks, or an unrelated baseline update.\n\n## Audit workflow\n\nAudit coherent owner batches, not deletion counts. Before deleting or\nconsolidating proof:\n\n1. Read the full test, production owner, callers, fixture and relevant history.\n2. Identify the observable contract and every distinct remaining proof layer.\n3. Confirm the candidate is duplicate, implementation-coupled, obsolete, or\n   unable to fail for the claimed regression.\n4. Preserve independent protocol, persistence, packaging, security,\n   architecture, rendering and native-runtime contracts.\n5. Remove obsolete test-only production seams in the same authorized change.\n\nBaseline failures may identify product bugs. Do not delete or weaken the test\nuntil the failure is classified.\n\n## Cost and validation\n\nFor a bug fix, demonstrate the intended pre-fix failure and post-fix pass, or\nreport why baseline reproduction was unavailable. For a refactor, preserve\nthe contract without pinning incidental implementation names.\n\nRun focused owner tests, then the\n[required validation](..\u002F..\u002F..\u002FAGENTS.md#required-validation-after-every-change).\nRespect explicit skip policies: an allowed ordinary CI skip is not complete\ncustom proof. A seconds-long test must protect a boundary that a cheaper layer\ncannot. Measure wall time when cost or performance is the claim, but do not\nimport JavaScript timing budgets, worker counts, or release-only rules into\nnative Windows suites.\n\nReport protected behavior, why the owner\u002Flayer is necessary, fixture limits,\nexact commands, execution counts, measured cost when relevant, and remaining\nproof gaps. Do not claim tests ran from an exit code alone.\n\n## Source and maintenance\n\nWindows-owned adaptation of concepts from the MIT-licensed upstream\n[test-audit workflow](https:\u002F\u002Fgithub.com\u002Fopenclaw\u002Fopenclaw\u002Fblob\u002F690c46084d507ce38502d401ef5d0d6a2fb9733a\u002F.agents\u002Fskills\u002Ftest-audit\u002FSKILL.md)\nand [test-authoring guide](https:\u002F\u002Fgithub.com\u002Fopenclaw\u002Fopenclaw\u002Fblob\u002F690c46084d507ce38502d401ef5d0d6a2fb9733a\u002Fdocs\u002Fhelp\u002Ftesting\u002Fwriting-tests.md).\nWindows retains its mandatory local closeout, native proof, and legitimate\nsource guards. This file does not inherit upstream publication permissions.\n","# Windows test audit\n\nUse when adding, changing, or reviewing tests. A review does not authorize\ntest deletion or production refactoring.\n\n## Authoring gate\n\nAnswer these questions:\n\n1. What observable behavior or independent contract does it protect?\n2. What credible regression makes it fail, and does it fail for that reason\n   rather than a broken fixture?\n3. Which production owner is exercised? Why does existing coverage miss this\n   risk?\n4. Are the expected values independent of the implementation under test?\n5. Does another layer add a distinct transport, persistence, lifecycle,\n   permission, or rendering risk, rather than repeat the same assertion?\n6. Is this the cheapest faithful boundary, or does it add a process, desktop,\n   WSL, installer, Gateway, or native capability boot that a smaller owner test\n   can cover?\n\nPrefer an existing table, fixture, or owner regression over another process\nboot. Avoid mocks supplying the behavior being asserted, production helpers\ncomputing their own expected values, and test-only exports without a genuine\nproduction boundary.\n\nRead [regression authoring](..\u002F..\u002F..\u002Fdocs\u002FTEST_COVERAGE.md#writing-a-regression-test)\nand use the existing [fixture guide](..\u002F..\u002F..\u002Fdocs\u002FGATEWAY_FIXTURE_TESTING.md)\nwhen the behavior crosses the application boundary.\n\n## Choose the production owner\n\nKeep one primary regression at the strongest faithful owner boundary:\n\n| Boundary | Primary project or lane |\n| --- | --- |\n| Gateway registry, credentials, connection lifecycle, migration admission\u002Frecords and persisted Local AI runtime state | `OpenClaw.Connection.Tests` |\n| Protocol, MCP, capabilities, command policy and shared models | `OpenClaw.Shared.Tests` |\n| Tray policy, settings, onboarding helpers, installer source contracts and update checks\u002Fbackends | `OpenClaw.Tray.Tests` |\n| Setup engine, WSL installation and Local AI setup\u002Frecovery policy | `OpenClaw.SetupEngine.Tests` |\n| Shipped `winnode` arguments, JSON output and command documentation | `OpenClaw.WinNode.Cli.Tests` |\n| Functional UI element graphs, factories and hook state without mounted native controls | `OpenClawTray.FunctionalUI.Tests` |\n| Mounted native WinUI controls, UI Automation, accessibility and focused input | `OpenClaw.Tray.UITests` |\n| Clean install, upgrade, uninstall and Store migration runtime behavior | The applicable formal validation path or proof pool |\n| Real process, Gateway-mediated setup\u002Fconnect or MXC behavior | Integration, E2E or named proof lane |\n\nAdd a second layer only when it proves a different failure mode. Local MCP\noutput does not prove Gateway pairing. A state snapshot is not a rendered UI\nassertion. Synthetic Gateway proof is not live provider, MXC, installer, Store,\nor production credential-resolution proof.\n\n## Reject low-value proof\n\nReject or rewrite tests that rely on:\n\n- self-comparisons, copied inventories, subject-generated expected values, or\n  the same helper on both sides of an assertion;\n- implementation names, private call shape, or exact delegation counts unless\n  that ownership\u002Fdelegation boundary is itself the protected contract;\n- a mock or fake that supplies the result, ordering, persistence receipt, or\n  policy decision being asserted;\n- another test at every crossed layer without a distinct transport,\n  persistence, permission, lifecycle, packaging, or rendering risk;\n- production flags or hooks that bypass owner logic, or exports\u002Fvisibility\n  added only to expose internal behavior without a genuine production boundary;\n- a broad source grep that pins incidental identifiers instead of a concrete\n  prohibited pattern or independent shipped inventory.\n\nNegative controls must fail at the intended guard, not during fixture setup.\nTest names must describe the behavior actually exercised. If a production seam\nexists only for obsolete tests and has no production-boundary or architecture\nrole, remove it with the tests when the task authorizes production changes;\notherwise report it rather than expanding it. Do not reject a narrow adapter\nover process, WSL, OS, network, clock, or filesystem boundaries merely because\ntests supply its only alternate implementation. Existing test assembly\nvisibility and ledger-recorded seams remain governed by their real contracts\nand retirement conditions.\n\n## Timing and determinism checklist\n\nFor native UI, loopback, process, and fixture tests, check each item against the\n[timing rules](..\u002F..\u002F..\u002Fdocs\u002FTEST_COVERAGE.md#deterministic-native-and-timing-sensitive-tests):\n\n- Readiness waits on an observable condition, not a delay, sleep, or fixed\n  settle count.\n- UI Automation elements are acquired inside the wait and used from that\n  capture, not looked up once with `!` or a throwing helper.\n- Geometry is asserted only after bounds stop moving, with an explicit\n  tolerance.\n- A polling predicate keeps polling while ownership is unproven and fails\n  fast only on positive foreign evidence.\n- Fixture data sets every value the assertion depends on; ports and other\n  inputs are deterministic.\n- Hang guards are named, environment-overridable constants with headroom for\n  a loaded runner. Latency-bounded work uses a dedicated thread, not the shared\n  thread pool.\n- A surviving flake is investigated as a possible product race before the\n  wait is changed.\n\nThe `TestTimingPatternGuard` ratchet rejects new literal short hang guards,\nfixed delays, and one-shot UI Automation lookups. Do not raise its baseline to\nadmit new code; use the named helper or a justified\n`\u002F\u002F test-timing-exception: \u003Creason>` comment. A rename or move may carry an\nexisting baseline entry to the new file when the rule totals stay unchanged.\nWhen a change removes occurrences, lower the baseline in the same change with\n`.\\scripts\\validate-test-timing-patterns.ps1 -LowerBaseline`.\n\n## Preserve valuable contracts\n\n- Keep independent protocol, CLI, storage, migration, defaults, packaging,\n  update, security, architecture, generated-resource, and user-visible\n  ordering contracts.\n- Static does not mean disposable. Source-shape guards are appropriate for a\n  concrete prohibited pattern with the\n  [ledger's retirement condition](..\u002F..\u002F..\u002Fdocs\u002FARCHITECTURE.md#how-to-use-this-document).\n  Installer source assertions remain distinct from clean install, upgrade, and\n  uninstall proof. The shipped `winnode` command-to-registry contract must\n  remain covered.\n- Use isolated settings, loopback listeners, owned processes, and existing\n  `TempDirectory`, `EnvironmentScope`, `CliHarness`, and Gateway fixture\n  helpers. Fakes may provide external input or record calls, but must not\n  implement the behavior under test.\n- Do not introduce arbitrary sleeps, blind polling, redundant process boots,\n  or parallel native input tests on a shared desktop. Reuse suite-level\n  fixtures when they preserve the same contract and isolation. Use a fresh\n  process when startup, process-global state, ownership\u002Ftermination, or an\n  independent profile is the boundary under test.\n- Do not make a failure green through weaker assertions, hidden skips,\n  blind retries, broadened mocks, or an unrelated baseline update.\n\n## Audit workflow\n\nAudit coherent owner batches, not deletion counts. Before deleting or\nconsolidating proof:\n\n1. Read the full test, production owner, callers, fixture and relevant history.\n2. Identify the observable contract and every distinct remaining proof layer.\n3. Confirm the candidate is duplicate, implementation-coupled, obsolete, or\n   unable to fail for the claimed regression.\n4. Preserve independent protocol, persistence, packaging, security,\n   architecture, rendering and native-runtime contracts.\n5. Remove obsolete test-only production seams in the same authorized change.\n\nBaseline failures may identify product bugs. Do not delete or weaken the test\nuntil the failure is classified.\n\n## Cost and validation\n\nFor a bug fix, demonstrate the intended pre-fix failure and post-fix pass, or\nreport why baseline reproduction was unavailable. For a refactor, preserve\nthe contract without pinning incidental implementation names.\n\nRun focused owner tests, then the\n[required validation](..\u002F..\u002F..\u002FAGENTS.md#required-validation-after-every-change).\nRespect explicit skip policies: an allowed ordinary CI skip is not complete\ncustom proof. A seconds-long test must protect a boundary that a cheaper layer\ncannot. Measure wall time when cost or performance is the claim, but do not\nimport JavaScript timing budgets, worker counts, or release-only rules into\nnative Windows suites.\n\nReport protected behavior, why the owner\u002Flayer is necessary, fixture limits,\nexact commands, execution counts, measured cost when relevant, and remaining\nproof gaps. Do not claim tests ran from an exit code alone.\n\n## Source and maintenance\n\nWindows-owned adaptation of concepts from the MIT-licensed upstream\n[test-audit workflow](https:\u002F\u002Fgithub.com\u002Fopenclaw\u002Fopenclaw\u002Fblob\u002F690c46084d507ce38502d401ef5d0d6a2fb9733a\u002F.agents\u002Fskills\u002Ftest-audit\u002FSKILL.md)\nand [test-authoring guide](https:\u002F\u002Fgithub.com\u002Fopenclaw\u002Fopenclaw\u002Fblob\u002F690c46084d507ce38502d401ef5d0d6a2fb9733a\u002Fdocs\u002Fhelp\u002Ftesting\u002Fwriting-tests.md).\nWindows retains its mandatory local closeout, native proof, and legitimate\nsource guards. This file does not inherit upstream publication permissions.\n",{"name":5,"description":7},"main","e33f70dfe31124b1acf81d42c7042a01d8c35e20",".agents\u002Fskills\u002Ftest-audit",349,"https:\u002F\u002Fgithub.com\u002Fopenclaw\u002Fopenclaw-windows-node","https:\u002F\u002Fopenclaw.ai","https:\u002F\u002Fgithub.com\u002Fopenclaw","Your personal, open source AI assistant.","2026-11-14 18:13:49",{"source":8,"skillName":5,"commands":53,"prompt":63,"zipAvailable":64},[54,57,60],{"tool":55,"command":56},"npx","npx skills add openclaw\u002Fopenclaw-windows-node --skill test-audit",{"tool":58,"command":59},"bunx","bunx skills add openclaw\u002Fopenclaw-windows-node --skill test-audit",{"tool":61,"command":62},"pnpm","pnpm dlx skills add openclaw\u002Fopenclaw-windows-node --skill test-audit","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fopenclaw\u002Fopenclaw-windows-node\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Ftest-audit\u002FSKILL.md",true,[],[67,75,100,116,133,149],{"skillNo":68,"name":69,"nameZh":6,"description":70,"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":74},"01M7416EHCYH7YCJ9XPSQ2844V","network-handoff","Use before generating, changing, or reviewing networking, IPC, or cross-process handoffs, including requests, retries, deadlines, reconnect, process launch, configuration apply, cancellation, and cleanup. Establish confirmation, identity, authority, reconciliation, and producer compatibility before implementation.","https:\u002F\u002Fgithub.com\u002Fopenclaw\u002Fopenclaw-windows-node\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fnetwork-handoff\u002FSKILL.md",[],[],"2026-11-11 15:11:14",{"skillNo":76,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":38},"01M6KZCXK5PCYWWW5FCVDYPAXK","retrigger-flaky-ci","Triage failed CI on a pull request or the active branch, decide whether the failures come from its changes, and rerun only failures that are unrelated. Use when asked to retrigger, rerun, or kick CI on a PR with failing checks.","https:\u002F\u002Fgithub.com\u002Fopenclaw\u002Fopenclaw-windows-node\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fretrigger-flaky-ci\u002FSKILL.md",[],[82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99],"retrigger flaky ci","ci","when","pull request","failing","rerun","use","triage","only","decide","request","active","pr","changes","unrelated","branch","with","pull",{"skillNo":101,"name":102,"nameZh":6,"description":103,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":104,"categories":105,"keywords":106,"publishedAt":38},"01M698QMQN5TQES4R7ZPTYAZF5","technical-documentation","Write, review, or restructure Windows node documentation and agent instructions using current source, task-focused explanations, and explicit fact preservation.","https:\u002F\u002Fgithub.com\u002Fopenclaw\u002Fopenclaw-windows-node\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Ftechnical-documentation\u002FSKILL.md",[],[107,108,109,110,111,32,35,112,27,113,114,115],"technical","documentation","technical documentation","node","instructions","task","agent instructions","preservation","source",{"skillNo":117,"name":118,"nameZh":6,"description":119,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":120,"categories":121,"keywords":122,"publishedAt":38},"01M698QKPXJ8ZG34MYHPPCAT4T","pr-simplify","Simplify your own Windows-node change to its smallest complete fix while preserving distinct regressions, platform contracts, and required proof. Use after implementation before PR preparation, after large repairs, or when asked to shrink a change or remove slop.","https:\u002F\u002Fgithub.com\u002Fopenclaw\u002Fopenclaw-windows-node\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fpr-simplify\u002FSKILL.md",[],[123,94,124,125,37,36,126,27,127,84,88,128,129,130,131,132,110],"pr simplify","simplify","preparation","implementation","change","fix","complete","remove","slop","platform",{"skillNo":134,"name":135,"nameZh":6,"description":136,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":137,"categories":138,"keywords":140,"publishedAt":38},"01M698QJEA384YPM8R9CPZ1YFD","openclaw-testing","Choose proportional OpenClaw Windows node tests, builds, packaging checks, and runtime proof, then diagnose failures without weakening required gates.","https:\u002F\u002Fgithub.com\u002Fopenclaw\u002Fopenclaw-windows-node\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fopenclaw-testing\u002FSKILL.md",[139],{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},[9,141,142,143,27,144,36,110,145,146,30,147,148],"openclaw testing","testing","diagnose","runtime","gates","runtime proof","choose","packaging",{"skillNo":150,"name":151,"nameZh":6,"description":152,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":153,"categories":154,"keywords":158,"publishedAt":38},"01M698QH7JVVA09769ER9P2TKY","openclaw-proof-validation","Plan and collect OpenClaw Windows validation\u002Fproof: tests, rubber-duck review, UI evidence, MCP output, and gateway runtime proof.","https:\u002F\u002Fgithub.com\u002Fopenclaw\u002Fopenclaw-windows-node\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fopenclaw-proof-validation\u002FSKILL.md",[155],{"categoryNo":156,"categoryName":157,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000023","Skill与智能体开发",[36,159,160,9,146,161,162,163,164,165,32,30,166,144,167,27,168,169],"validation","openclaw proof validation","duck","ui","ui evidence","plan","output","collect","evidence","gateway","MCP"]