[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6DBTH02Z4MCF1Z6JT8VVGNF":3,"skill-same:01M6DBTH02Z4MCF1Z6JT8VVGNF":58},{"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":33,"linkOnly":34,"markdownContent":35,"markdownBody":36,"frontMatter":37,"refName":38,"commitSha":39,"skillDirectory":40,"forksCount":14,"repositoryHtmlUrl":41,"repositoryHomepageUrl":6,"ownerHtmlUrl":42,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":43,"install":44,"resources":57},"01M6DBTH02Z4MCF1Z6JT8VVGNF","rtv-pdr",null,"Use when a hard coding work-unit has failed once, is contentious, or scores HIGH on the ESE (Ensemble Semantic Entropy) gate — spawns N parallel Opus rollouts, runs a recursive tournament where fresh judge agents (never contestants) pick the winner, then optionally runs a PDR refine pass. Expensive; do NOT fire on routine tasks.","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\u002F.agents\u002Fskills\u002Frtv-pdr\u002FSKILL.md",[],[19,20,21,22,23,24,25,26,27,28,29,30,31,32],"rtv pdr","tasks","entropy","coding","parallel","pick","work","use","never","rollouts","gate","do","refine","fresh","2026-11-10 10:12:30",false,"---\nname: rtv-pdr\ndescription: Use when a hard coding work-unit has failed once, is contentious, or scores HIGH on the ESE (Ensemble Semantic Entropy) gate — spawns N parallel Opus rollouts, runs a recursive tournament where fresh judge agents (never contestants) pick the winner, then optionally runs a PDR refine pass. Expensive; do NOT fire on routine tasks.\n---\n\n# RTV+PDR — Recursive Tournament Voting + Parallel-Distill-Refine\n\nTest-time-compute scaling for hard coding work-units.\nBased on the methodology in arXiv:2604.16529.\n\n**Research caveat:** The cited paper reports gains for Claude Opus 4.5 on SWE-bench Verified (70.9→77.6) and Terminal-Bench v2.0 (46.9→59.1). Results are model-version-specific and benchmark-specific; gains on arbitrary user tasks are not guaranteed.\n\n## When to use\n\nFire this workflow **only** when all of the following hold:\n\n- The work-unit is **hard or contentious** (prior attempt failed, or high disagreement among reviewers).\n- The ESE (Ensemble Semantic Entropy) escalation gate has **flagged** the unit as high-entropy.\n- You have **budget headroom** — each invocation runs N Opus-class agents (rollouts + judges + optional refine).\n\nDo NOT use for routine implementation, boilerplate, or tasks with an obvious single solution.\n\n## Cost profile\n\n| N | Rollout calls | Tournament judges (worst case) | PDR refine | Total (approx) |\n|---|---|---|---|---|\n| 3 | 3 Opus | 2 Opus | 1 Opus + 1 judge | 7 Opus |\n| 4 | 4 Opus | 3 Opus | 1 Opus + 1 judge | 9 Opus |\n| 5 | 5 Opus | 4 Opus | 1 Opus + 1 judge | 11 Opus |\n\n`CLAUDE_CODE_FORK_SUBAGENT` fork-cache is used automatically for parallel rollouts, cutting warmup cost.\n\nJudges use `model: opus` by default (best verdict quality). To reduce cost, you may edit the judge calls to `model: 'sonnet'`; document this tradeoff if you do.\n\n## Judge independence guarantee\n\n**Judges are never contestants.** Each pairwise comparison is run by a fresh agent that received no prior context from either rollout. This closes the self-choice bias gap identified in the RTV design.\n\n`self_reported_confidence` in each rollout summary is logged for observability only — **it is never used to select the winner**.\n\n## Invocation\n\nRun the workflow from the Claude Code Workflow runner by selecting `rtv-pdr` and passing args:\n\n```\nargs:\n  task: \"\u003Cfull description of the hard coding work-unit>\"\n  repo_context: \"\u003Crelevant repo context, file paths, constraints>\"\n  n: 4          # number of parallel rollouts (2-5, default 4)\n  pdr_mode: false  # set true to run a PDR refine pass after the tournament\n```\n\n### ESE-gated invocation pattern\n\nOnly invoke after ESE confirms high semantic entropy:\n\n```\n1. Run ESE gate on the work-unit.\n2. If ESE verdict == HIGH_ENTROPY:\n     Invoke rtv-pdr workflow with args.task, args.repo_context, args.n=4\n3. If ESE verdict == LOW_ENTROPY:\n     Use standard single-agent implementation (do not pay RTV cost).\n```\n\n### PDR mode\n\nSet `args.pdr_mode: true` when a previous attempt has already failed and you have failure context to exploit. PDR runs one additional refine rollout conditioned on:\n- The tournament winner's distilled summary\n- All losers' `failure_modes` (structured, not raw history)\n\nThe refine result is then judged against the tournament winner by a fresh agent. Use sparingly — it adds ~2 Opus calls.\n\n## Output shape\n\n```json\n{\n  \"winner\": { \"id\": \"R2\", \"summary\": { ...ROLLOUT_SCHEMA... } },\n  \"runner_up\": { \"id\": \"R0\", \"summary\": { ...ROLLOUT_SCHEMA... } },\n  \"rounds\": 2,\n  \"rationale\": \"Winner R2 selected by tournament after 2 round(s). ...\"\n}\n```\n\n`runner_up` is the finalist that lost the final match (`null` if fewer than 2 rollouts succeeded).\n\n## Workflow location\n\n`workflows\u002Frtv-pdr.js` in this toolchain repo.\n","# RTV+PDR — Recursive Tournament Voting + Parallel-Distill-Refine\n\nTest-time-compute scaling for hard coding work-units.\nBased on the methodology in arXiv:2604.16529.\n\n**Research caveat:** The cited paper reports gains for Claude Opus 4.5 on SWE-bench Verified (70.9→77.6) and Terminal-Bench v2.0 (46.9→59.1). Results are model-version-specific and benchmark-specific; gains on arbitrary user tasks are not guaranteed.\n\n## When to use\n\nFire this workflow **only** when all of the following hold:\n\n- The work-unit is **hard or contentious** (prior attempt failed, or high disagreement among reviewers).\n- The ESE (Ensemble Semantic Entropy) escalation gate has **flagged** the unit as high-entropy.\n- You have **budget headroom** — each invocation runs N Opus-class agents (rollouts + judges + optional refine).\n\nDo NOT use for routine implementation, boilerplate, or tasks with an obvious single solution.\n\n## Cost profile\n\n| N | Rollout calls | Tournament judges (worst case) | PDR refine | Total (approx) |\n|---|---|---|---|---|\n| 3 | 3 Opus | 2 Opus | 1 Opus + 1 judge | 7 Opus |\n| 4 | 4 Opus | 3 Opus | 1 Opus + 1 judge | 9 Opus |\n| 5 | 5 Opus | 4 Opus | 1 Opus + 1 judge | 11 Opus |\n\n`CLAUDE_CODE_FORK_SUBAGENT` fork-cache is used automatically for parallel rollouts, cutting warmup cost.\n\nJudges use `model: opus` by default (best verdict quality). To reduce cost, you may edit the judge calls to `model: 'sonnet'`; document this tradeoff if you do.\n\n## Judge independence guarantee\n\n**Judges are never contestants.** Each pairwise comparison is run by a fresh agent that received no prior context from either rollout. This closes the self-choice bias gap identified in the RTV design.\n\n`self_reported_confidence` in each rollout summary is logged for observability only — **it is never used to select the winner**.\n\n## Invocation\n\nRun the workflow from the Claude Code Workflow runner by selecting `rtv-pdr` and passing args:\n\n```\nargs:\n  task: \"\u003Cfull description of the hard coding work-unit>\"\n  repo_context: \"\u003Crelevant repo context, file paths, constraints>\"\n  n: 4          # number of parallel rollouts (2-5, default 4)\n  pdr_mode: false  # set true to run a PDR refine pass after the tournament\n```\n\n### ESE-gated invocation pattern\n\nOnly invoke after ESE confirms high semantic entropy:\n\n```\n1. Run ESE gate on the work-unit.\n2. If ESE verdict == HIGH_ENTROPY:\n     Invoke rtv-pdr workflow with args.task, args.repo_context, args.n=4\n3. If ESE verdict == LOW_ENTROPY:\n     Use standard single-agent implementation (do not pay RTV cost).\n```\n\n### PDR mode\n\nSet `args.pdr_mode: true` when a previous attempt has already failed and you have failure context to exploit. PDR runs one additional refine rollout conditioned on:\n- The tournament winner's distilled summary\n- All losers' `failure_modes` (structured, not raw history)\n\nThe refine result is then judged against the tournament winner by a fresh agent. Use sparingly — it adds ~2 Opus calls.\n\n## Output shape\n\n```json\n{\n  \"winner\": { \"id\": \"R2\", \"summary\": { ...ROLLOUT_SCHEMA... } },\n  \"runner_up\": { \"id\": \"R0\", \"summary\": { ...ROLLOUT_SCHEMA... } },\n  \"rounds\": 2,\n  \"rationale\": \"Winner R2 selected by tournament after 2 round(s). ...\"\n}\n```\n\n`runner_up` is the finalist that lost the final match (`null` if fewer than 2 rollouts succeeded).\n\n## Workflow location\n\n`workflows\u002Frtv-pdr.js` in this toolchain repo.\n",{"name":5,"description":7},"main","a298c929b6cbd662e1fce54302ce520d5ceea1a5",".agents\u002Fskills\u002Frtv-pdr","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain","https:\u002F\u002Fgithub.com\u002FPrekzursil","2026-11-02 14:30:57",{"source":8,"skillName":5,"commands":45,"prompt":55,"zipAvailable":56},[46,49,52],{"tool":47,"command":48},"npx","npx skills add Prekzursil\u002Fagent-skills-toolchain --skill rtv-pdr",{"tool":50,"command":51},"bunx","bunx skills add Prekzursil\u002Fagent-skills-toolchain --skill rtv-pdr",{"tool":53,"command":54},"pnpm","pnpm dlx skills add Prekzursil\u002Fagent-skills-toolchain --skill rtv-pdr","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Frtv-pdr\u002FSKILL.md",true,[],[59,80,109,136,150,165],{"skillNo":60,"name":61,"nameZh":6,"description":62,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":63,"categories":64,"keywords":65,"publishedAt":33},"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",[],[66,67,68,69,70,71,72,73,74,26,75,76,77,78,79],"clarify requirements","clarify","requirements","risk","options","task","prompts","goal","scanning","context","repo","injection","execution","file",{"skillNo":81,"name":82,"nameZh":6,"description":83,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":84,"categories":85,"keywords":91,"publishedAt":33},"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",[86],{"categoryNo":87,"categoryName":88,"taxonomyType":89,"taxonomyTypeDesc":90},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[92,93,94,24,95,96,97,98,99,100,101,102,103,104,26,105,106,107,30,108],"openmm md","how","decision","linux","analyze","tool","stack","open","NVIDIA","gpu","license","workflow","windows","free","molecular dynamics","local","MCP",{"skillNo":110,"name":111,"nameZh":6,"description":112,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":113,"categories":114,"keywords":121,"publishedAt":33},"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",[115,118],{"categoryNo":116,"categoryName":117,"taxonomyType":89,"taxonomyTypeDesc":90},"01JBSEEDCAT000000000000003","演示文稿",{"categoryNo":119,"categoryName":120,"taxonomyType":89,"taxonomyTypeDesc":90},"01JBSEEDCAT000000000000008","测试与质量",[122,123,124,125,126,127,128,129,130,131,93,132,133,134,135],"review","visual","visual review","pages","web","false","iteration","take","check","slides","ui","screenshot","playwright","screenshots",{"skillNo":137,"name":138,"nameZh":6,"description":139,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":140,"categories":141,"keywords":142,"publishedAt":33},"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",[],[138,143,144,145,146,147,148,149],"setup","report","project","installation","project setup","issues","state",{"skillNo":151,"name":152,"nameZh":6,"description":153,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":154,"categories":155,"keywords":157,"publishedAt":33},"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",[156],{"categoryNo":87,"categoryName":88,"taxonomyType":89,"taxonomyTypeDesc":90},[152,158,71,159,160,161,26,162,25,163,164],"start issue","issue","ready","work on issue","user","work on","development",{"skillNo":166,"name":143,"nameZh":6,"description":167,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":168,"categories":169,"keywords":170,"publishedAt":33},"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",[],[143,145,107,171,147,172],"files","interactive"]