[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6DF17B5PE3C5GZPYNH14EXT":3,"skill-same:01M6DF17B5PE3C5GZPYNH14EXT":70},{"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":39,"linkOnly":40,"markdownContent":41,"markdownBody":42,"frontMatter":43,"refName":44,"commitSha":45,"skillDirectory":46,"forksCount":14,"repositoryHtmlUrl":47,"repositoryHomepageUrl":6,"ownerHtmlUrl":48,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":49,"install":50,"resources":63},"01M6DF17B5PE3C5GZPYNH14EXT","triage-dispatch",null,"Use when repairing an app with MULTIPLE bugs or broken user-journeys at once (not a single fix) - instead of edit-one-file \u002F rerun-the-whole-suite \u002F hope, first map the broken journeys with evidence, capture each REAL error, build a defect graph, then dispatch parallel bounded-scope fix agents and verify only impacted regions before an outer full+e2e+mutation gate.","Prekzursil\u002Fagent-skills-toolchain","Prekzursil","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F54636077?v=4","01M6DACZ1PCQHD99SCR1YXG28B","Andrei Mihai Visalon","01M6DACZ670WZF535V3WYYSS4Z",0,"MIT","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Fskills\u002Ftriage-dispatch\u002FSKILL.md",[],[19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38],"triage dispatch","dispatch","triage","use","app","mutation","map","build","gate","evidence","only","fix","edit","broken","capture","parallel","verify","first","e2e","file","2026-11-10 10:12:30",false,"---\nname: triage-dispatch\ndescription: Use when repairing an app with MULTIPLE bugs or broken user-journeys at once (not a single fix) - instead of edit-one-file \u002F rerun-the-whole-suite \u002F hope, first map the broken journeys with evidence, capture each REAL error, build a defect graph, then dispatch parallel bounded-scope fix agents and verify only impacted regions before an outer full+e2e+mutation gate.\n---\n\n# triage-dispatch — audit → defect-graph → parallel bounded fixes → verify\n\nThe repair method that beats blind edit-one-file-then-rerun-the-whole-suite. It exists because\nscattershot repair wastes tokens and expands SCOPE instead of precision (ProjDevBench ~138 turns \u002F\n4.81M tokens per end-to-end problem; the 20,574-session developer-agent study: agent edits grow the\nchange footprint, not the fix precision). Pair it with `two-tier-verify`, `seed-golden-journey`, and\n`acceptance-review`.\n\n## When to use\nMultiple bugs \u002F several broken journeys \u002F a \"held together with chewing gum\" app where you'd otherwise\nfix one thing, re-run everything, and chase symptoms. Skip it for a single, well-understood one-liner.\n\n## The method\n\n### 1. Audit-first — map the broken journeys with evidence\nEnumerate the app's real user journeys and, for each, whether it works — with EVIDENCE (a run, a\nproduced artifact, a screenshot), not a guess. The output is a list of BROKEN journeys, each with a\nconcrete failing observation. Coverage\u002Fgreen units are not evidence here.\n\n### 2. Capture the REAL error — never infer the stage from artifacts\nFor each broken journey, capture the actual error before touching code. A surviving artifact on disk\ntells you which stage RAN, not which stage FAILED. If the failure is in a background job \u002F sidecar whose\nerror is not surfaced, drive it directly and drain its stderr (e.g. spawn the sidecar over its JSON-RPC,\nsend the request, read notifications to done, collect the logged traceback). Inferring the stage costs\nfix-the-wrong-thing detours.\n\n### 3. Build the defect graph\nTrace each symptom to its code site(s) and record dependencies between fixes (which must land before\nwhich; which share a file; which are independent). A defect graph localizes work and reveals safe\nparallelism (GraphLocator, FSE'26: +function-recall \u002F +resolve-rate over blind localization). Note the\ncommon-root clusters — several symptoms often share one cause (a CRLF line-ending, a Windows-path escape,\na teardown crash).\n\n### 4. Dispatch parallel fix agents — bounded scope, isolated\nFor independent defect-subgraphs, dispatch one fix agent each, with a DECLARED file scope, in an isolated\nworktree (so parallel edits can't collide). Re-anchor the original goal + the specific defect at the top\nof each handoff (structured summary, never raw prior output). A bounded scope is the guard against the\nscope-creep the misalignment study measured.\n\n### 5. Inner-loop verify only impacted regions\nPer fix, run only the affected\u002Frelated tests + type + lint on touched files (see `two-tier-verify` inner\nloop) — Ekstazi\u002FTIA\u002FCloudBees cut test time 32-70%. Do not re-run the whole suite per edit.\n\n### 6. Outer gate at the commit boundary\nOnce fixes land: full suite + coverage + the held-out golden-journey\u002FE2E + `mutation-check`, ONCE. On an\nouter failure, map it back to the defect graph (git bisect \u002F the one failing test) rather than re-auditing.\n\n### 7. Independent acceptance review\nBefore declaring done, run `acceptance-review`: a fresh-context \u002F cross-model reviewer over the diff +\nthe journeys + the real artifacts. SHIP only if every declared journey has a real post-change artifact.\nNever self-certify.\n\n## Anti-patterns this replaces\n- Edit one file → rerun the entire suite → repeat (token churn, no localization).\n- Inferring the failing stage from what landed on disk (fix-the-wrong-thing).\n- One agent grinding all bugs serially in one context (scope creep, drift, degraded verification).\n- \"Green units \u002F 100% coverage\" cited as done (Goodhart — see `seed-golden-journey`).\n\n## Cross-links\n`seed-golden-journey` (the done-signal these fixes must satisfy) · `two-tier-verify` (the inner\u002Fouter\nloops) · `acceptance-review` (the independent verifier) · `mutation-check` (test-strength at the gate) ·\n`fresh-context-verify` (re-derive a load-bearing anchor in a clean context).\n","# triage-dispatch — audit → defect-graph → parallel bounded fixes → verify\n\nThe repair method that beats blind edit-one-file-then-rerun-the-whole-suite. It exists because\nscattershot repair wastes tokens and expands SCOPE instead of precision (ProjDevBench ~138 turns \u002F\n4.81M tokens per end-to-end problem; the 20,574-session developer-agent study: agent edits grow the\nchange footprint, not the fix precision). Pair it with `two-tier-verify`, `seed-golden-journey`, and\n`acceptance-review`.\n\n## When to use\nMultiple bugs \u002F several broken journeys \u002F a \"held together with chewing gum\" app where you'd otherwise\nfix one thing, re-run everything, and chase symptoms. Skip it for a single, well-understood one-liner.\n\n## The method\n\n### 1. Audit-first — map the broken journeys with evidence\nEnumerate the app's real user journeys and, for each, whether it works — with EVIDENCE (a run, a\nproduced artifact, a screenshot), not a guess. The output is a list of BROKEN journeys, each with a\nconcrete failing observation. Coverage\u002Fgreen units are not evidence here.\n\n### 2. Capture the REAL error — never infer the stage from artifacts\nFor each broken journey, capture the actual error before touching code. A surviving artifact on disk\ntells you which stage RAN, not which stage FAILED. If the failure is in a background job \u002F sidecar whose\nerror is not surfaced, drive it directly and drain its stderr (e.g. spawn the sidecar over its JSON-RPC,\nsend the request, read notifications to done, collect the logged traceback). Inferring the stage costs\nfix-the-wrong-thing detours.\n\n### 3. Build the defect graph\nTrace each symptom to its code site(s) and record dependencies between fixes (which must land before\nwhich; which share a file; which are independent). A defect graph localizes work and reveals safe\nparallelism (GraphLocator, FSE'26: +function-recall \u002F +resolve-rate over blind localization). Note the\ncommon-root clusters — several symptoms often share one cause (a CRLF line-ending, a Windows-path escape,\na teardown crash).\n\n### 4. Dispatch parallel fix agents — bounded scope, isolated\nFor independent defect-subgraphs, dispatch one fix agent each, with a DECLARED file scope, in an isolated\nworktree (so parallel edits can't collide). Re-anchor the original goal + the specific defect at the top\nof each handoff (structured summary, never raw prior output). A bounded scope is the guard against the\nscope-creep the misalignment study measured.\n\n### 5. Inner-loop verify only impacted regions\nPer fix, run only the affected\u002Frelated tests + type + lint on touched files (see `two-tier-verify` inner\nloop) — Ekstazi\u002FTIA\u002FCloudBees cut test time 32-70%. Do not re-run the whole suite per edit.\n\n### 6. Outer gate at the commit boundary\nOnce fixes land: full suite + coverage + the held-out golden-journey\u002FE2E + `mutation-check`, ONCE. On an\nouter failure, map it back to the defect graph (git bisect \u002F the one failing test) rather than re-auditing.\n\n### 7. Independent acceptance review\nBefore declaring done, run `acceptance-review`: a fresh-context \u002F cross-model reviewer over the diff +\nthe journeys + the real artifacts. SHIP only if every declared journey has a real post-change artifact.\nNever self-certify.\n\n## Anti-patterns this replaces\n- Edit one file → rerun the entire suite → repeat (token churn, no localization).\n- Inferring the failing stage from what landed on disk (fix-the-wrong-thing).\n- One agent grinding all bugs serially in one context (scope creep, drift, degraded verification).\n- \"Green units \u002F 100% coverage\" cited as done (Goodhart — see `seed-golden-journey`).\n\n## Cross-links\n`seed-golden-journey` (the done-signal these fixes must satisfy) · `two-tier-verify` (the inner\u002Fouter\nloops) · `acceptance-review` (the independent verifier) · `mutation-check` (test-strength at the gate) ·\n`fresh-context-verify` (re-derive a load-bearing anchor in a clean context).\n",{"name":5,"description":7},"main","a298c929b6cbd662e1fce54302ce520d5ceea1a5","skills\u002Ftriage-dispatch","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain","https:\u002F\u002Fgithub.com\u002FPrekzursil","2026-11-02 14:30:57",{"source":8,"skillName":5,"commands":51,"prompt":61,"zipAvailable":62},[52,55,58],{"tool":53,"command":54},"npx","npx skills add Prekzursil\u002Fagent-skills-toolchain --skill triage-dispatch",{"tool":56,"command":57},"bunx","bunx skills add Prekzursil\u002Fagent-skills-toolchain --skill triage-dispatch",{"tool":59,"command":60},"pnpm","pnpm dlx skills add Prekzursil\u002Fagent-skills-toolchain --skill triage-dispatch","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Fskills\u002Ftriage-dispatch\u002FSKILL.md",true,[64],{"path":65,"resourceType":66,"resourceTypeDesc":67,"sizeBytes":68,"sourceHtmlUrl":69},"README.md","REFERENCE","参考文档",1080,"https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Fskills\u002Ftriage-dispatch\u002FREADME.md",[71,91,122,149,163,179],{"skillNo":72,"name":73,"nameZh":6,"description":74,"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":39},"01M6DF7D7478R55P46REYH4QAC","clarify-requirements","Resolve genuine pre-execution ambiguity by scanning repo context, generating 1-6 grounded options, and asking 1-3 focused questions via AskUserQuestion. NOT for mid-task use (injection risk); NOT when the goal is already unambiguous; use to disambiguate vague prompts before any file edits.","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Fworkspace-hooks\u002Fskills\u002Fclarify-requirements\u002FSKILL.md",[],[78,79,80,81,82,83,84,85,86,22,87,88,89,90,38],"clarify requirements","clarify","requirements","risk","options","task","prompts","goal","scanning","context","repo","injection","execution",{"skillNo":92,"name":93,"nameZh":6,"description":94,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":95,"categories":96,"keywords":102,"publishedAt":39},"01M6DF79CA4BHJS66J8EZDBPWJ","openmm-md","Run FREE, open-source, Windows-native GPU molecular dynamics via the openmm-md MCP (OpenMM + CUDA on the local NVIDIA GPU). Use for protein \u002F protein-ligand MD when you want a free stack or are on Windows without a Schrodinger license. Covers the full workflow (prepare, solvate, minimize, equilibrate, production, analyze, mmgbsa) AND an honest decision guide vs Schrodinger Desmond, including how to run Desmond legitimately on Linux or a cloud GPU with a real license. This is a deliberately LESSER free alternative and does NOT replace the schrodinger MCP tools. Do NOT use it to pick a free tool when a validated FEP+, WaterMap, IFD-MD, or Prime MM-GBSA result is actually required.","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Fmcp\u002Fopenmm-md-mcp\u002Fskill\u002Fopenmm-md\u002FSKILL.md",[97],{"categoryNo":98,"categoryName":99,"taxonomyType":100,"taxonomyTypeDesc":101},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[103,104,105,106,107,108,109,110,111,112,113,114,115,116,22,117,118,119,120,121],"openmm md","how","decision","pick","linux","analyze","tool","stack","open","NVIDIA","gpu","license","workflow","windows","free","molecular dynamics","local","do","MCP",{"skillNo":123,"name":124,"nameZh":6,"description":125,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":126,"categories":127,"keywords":134,"publishedAt":39},"01M6DF76C1MG85VEWP8VZCBNMN","visual-review","Take screenshots of web pages and UI using Playwright for visual review and iteration","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Feditor\u002Fmetaswarm\u002Fskills\u002Fvisual-review\u002FSKILL.md",[128,131],{"categoryNo":129,"categoryName":130,"taxonomyType":100,"taxonomyTypeDesc":101},"01JBSEEDCAT000000000000003","演示文稿",{"categoryNo":132,"categoryName":133,"taxonomyType":100,"taxonomyTypeDesc":101},"01JBSEEDCAT000000000000008","测试与质量",[135,136,137,138,139,140,141,142,143,144,104,145,146,147,148],"review","visual","visual review","pages","web","false","iteration","take","check","slides","ui","screenshot","playwright","screenshots",{"skillNo":150,"name":151,"nameZh":6,"description":152,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":153,"categories":154,"keywords":155,"publishedAt":39},"01M6DF73AQVCBZKEN0YXEX9FXP","status","Diagnostic status report — shows metaswarm installation state, project setup, and potential issues","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Feditor\u002Fmetaswarm\u002Fskills\u002Fstatus\u002FSKILL.md",[],[151,156,157,158,159,160,161,162],"setup","report","project","installation","project setup","issues","state",{"skillNo":164,"name":165,"nameZh":6,"description":166,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":167,"categories":168,"keywords":170,"publishedAt":39},"01M6DF6X6C2XB0P6GX76P7TTYV","start","Use when starting work on any task, when the user mentions metaswarm, or when the user wants to begin tracked development work","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Feditor\u002Fmetaswarm\u002Fskills\u002Fstart\u002FSKILL.md",[169],{"categoryNo":98,"categoryName":99,"taxonomyType":100,"taxonomyTypeDesc":101},[165,171,83,172,173,174,22,175,176,177,178],"start issue","issue","ready","work on issue","user","work","work on","development",{"skillNo":180,"name":156,"nameZh":6,"description":181,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":182,"categories":183,"keywords":184,"publishedAt":39},"01M6DF6QMJ9AJKFKHHKRT005PQ","Interactive project setup — detects your project, configures metaswarm, writes project-local files","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Feditor\u002Fmetaswarm\u002Fskills\u002Fsetup\u002FSKILL.md",[],[156,158,119,185,160,186],"files","interactive"]