[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M71RVJ4DZ0RDCV7FQG2VAJMQ":3,"skill-same:01M71RVJ4DZ0RDCV7FQG2VAJMQ":54},{"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":26,"linkOnly":27,"markdownContent":28,"markdownBody":29,"frontMatter":30,"refName":31,"commitSha":32,"skillDirectory":33,"forksCount":34,"repositoryHtmlUrl":35,"repositoryHomepageUrl":6,"ownerHtmlUrl":36,"ownerBlogUrl":37,"ownerDescription":38,"updatedAt":39,"install":40,"resources":53},"01M71RVJ4DZ0RDCV7FQG2VAJMQ","implement-evaluation-steps",null,"Use when defining or updating Eve evaluation steps.","scaling-group\u002Feve","scaling-group","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F181708022?v=4","01M71RTM8RQW6CX3M1934895JA","Scientific Computing and Intelligence Group","01M71RTM99S28ZTFK9BK17R4GC",66,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fscaling-group\u002Feve\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Fimplement-evaluation-steps\u002FSKILL.md",[],[19,20,21,22,23,24,25],"implement","Evaluation","steps","implement evaluation steps","eve","use","when","2026-11-10 13:02:02",false,"---\nname: implement-evaluation-steps\ndescription: \"Use when defining or updating Eve evaluation steps.\"\n---\n\n\n# Implement Evaluation Steps\n\n## Overview\n\nEvaluation is a first-class config tree: `configs\u002Feve\u002Fevaluation\u002F\u003Capp>.yaml` composed via the\nrun config's `defaults:` list (e.g. `- evaluation: circle_packing`). The file\ndeclares an ordered `evaluation.steps` pipeline plus an `evaluation.failure_score`. Every run\nconfig MUST compose an `evaluation:` group; there is no implicit fallback.\n\n`evaluation.steps` is an ordered, arbitrarily-long pipeline. Dispatch is by FORM, not a `kind`\nfield:\n- a bare `.sh` path string -> programmatic (deterministic) step\n- a `{prompt, immutable?}` mapping -> LLM judge step (a first-class worker: its own prompt +\n  budget, plus an OPTIONAL immutable rubric + score-check subagent). Omit `immutable` for a\n  lightweight judge that runs from its ENTRYPOINT alone.\n\n## Config form\n\n```yaml\n# @package _global_\nevaluation:\n  steps:\n    - configs\u002Feve\u002Fevaluation\u002F\u003Capp>\u002Fevaluation.sh        # programmatic step\n    # - { prompt: configs\u002Feve\u002Fevaluation\u002F\u003Capp>\u002Fprompt_assess,\n    #     immutable: configs\u002Feve\u002Fevaluation\u002F\u003Capp>\u002Fimmutable_assess }   # judge step\n  failure_score:\n    score: 0.0\n    summary: evaluation failed\n  seed_solver_score: null\n  seed_solver_skip_evaluation: false\n```\n\n## Runtime Inputs (shell steps)\n\nShell evaluators receive these environment variables:\n- `EVE_WORKSPACE_ROOT`: workspace root\n- `EVE_SOLVER_ROOT`: solver candidate repo root, that is `EVE_WORKSPACE_ROOT\u002Fsolver`\n- `EVE_EVAL_LOG_ROOT`: formal evaluation log root, that is `EVE_WORKSPACE_ROOT\u002Flogs\u002Fevaluate`\n\nUse these paths directly in the shell file instead of reconstructing ad hoc paths.\n\n## Requirements\n\n- All shell steps run from `EVE_WORKSPACE_ROOT`, not `EVE_SOLVER_ROOT` or the downstream repo.\n- Write final evaluation files to `EVE_EVAL_LOG_ROOT`. This is the only interface that will be extracted.\n- `EVE_EVAL_LOG_ROOT\u002Fscore.yaml` is required.\n\n`score.yaml` rules:\n- `score.yaml` must always exist.\n- The score payload may be any YAML PyTree.\n- If evaluation fails, write a minimal YAML payload such as:\n\n```yaml\nscore: 0.0\nsummary: evaluation failed\n```\n\nYou can store files of any format within the `EVE_EVAL_LOG_ROOT` directory, including multimedia\nlike images and videos. If your logging structure is complex, it is recommended to include a\n`README.md` inside `EVE_EVAL_LOG_ROOT` to document the file organization and contents.\n\n## Reference Example\n\ncircle_packing ships both forms:\n- Default single-step: `configs\u002Feve\u002Fevaluation\u002Fcircle_packing.yaml`\n  pointing at `configs\u002Feve\u002Fevaluation\u002Fcircle_packing\u002Fevaluation.sh`.\n- Judge pipeline: `configs\u002Feve\u002Fevaluation\u002Fcircle_packing.judge.yaml`\n  (`performance.sh` -> `assess` -> `aggregate`), with prompt\u002Fimmutable dirs under\n  `configs\u002Feve\u002Fevaluation\u002Fcircle_packing\u002F`.\n","# Implement Evaluation Steps\n\n## Overview\n\nEvaluation is a first-class config tree: `configs\u002Feve\u002Fevaluation\u002F\u003Capp>.yaml` composed via the\nrun config's `defaults:` list (e.g. `- evaluation: circle_packing`). The file\ndeclares an ordered `evaluation.steps` pipeline plus an `evaluation.failure_score`. Every run\nconfig MUST compose an `evaluation:` group; there is no implicit fallback.\n\n`evaluation.steps` is an ordered, arbitrarily-long pipeline. Dispatch is by FORM, not a `kind`\nfield:\n- a bare `.sh` path string -> programmatic (deterministic) step\n- a `{prompt, immutable?}` mapping -> LLM judge step (a first-class worker: its own prompt +\n  budget, plus an OPTIONAL immutable rubric + score-check subagent). Omit `immutable` for a\n  lightweight judge that runs from its ENTRYPOINT alone.\n\n## Config form\n\n```yaml\n# @package _global_\nevaluation:\n  steps:\n    - configs\u002Feve\u002Fevaluation\u002F\u003Capp>\u002Fevaluation.sh        # programmatic step\n    # - { prompt: configs\u002Feve\u002Fevaluation\u002F\u003Capp>\u002Fprompt_assess,\n    #     immutable: configs\u002Feve\u002Fevaluation\u002F\u003Capp>\u002Fimmutable_assess }   # judge step\n  failure_score:\n    score: 0.0\n    summary: evaluation failed\n  seed_solver_score: null\n  seed_solver_skip_evaluation: false\n```\n\n## Runtime Inputs (shell steps)\n\nShell evaluators receive these environment variables:\n- `EVE_WORKSPACE_ROOT`: workspace root\n- `EVE_SOLVER_ROOT`: solver candidate repo root, that is `EVE_WORKSPACE_ROOT\u002Fsolver`\n- `EVE_EVAL_LOG_ROOT`: formal evaluation log root, that is `EVE_WORKSPACE_ROOT\u002Flogs\u002Fevaluate`\n\nUse these paths directly in the shell file instead of reconstructing ad hoc paths.\n\n## Requirements\n\n- All shell steps run from `EVE_WORKSPACE_ROOT`, not `EVE_SOLVER_ROOT` or the downstream repo.\n- Write final evaluation files to `EVE_EVAL_LOG_ROOT`. This is the only interface that will be extracted.\n- `EVE_EVAL_LOG_ROOT\u002Fscore.yaml` is required.\n\n`score.yaml` rules:\n- `score.yaml` must always exist.\n- The score payload may be any YAML PyTree.\n- If evaluation fails, write a minimal YAML payload such as:\n\n```yaml\nscore: 0.0\nsummary: evaluation failed\n```\n\nYou can store files of any format within the `EVE_EVAL_LOG_ROOT` directory, including multimedia\nlike images and videos. If your logging structure is complex, it is recommended to include a\n`README.md` inside `EVE_EVAL_LOG_ROOT` to document the file organization and contents.\n\n## Reference Example\n\ncircle_packing ships both forms:\n- Default single-step: `configs\u002Feve\u002Fevaluation\u002Fcircle_packing.yaml`\n  pointing at `configs\u002Feve\u002Fevaluation\u002Fcircle_packing\u002Fevaluation.sh`.\n- Judge pipeline: `configs\u002Feve\u002Fevaluation\u002Fcircle_packing.judge.yaml`\n  (`performance.sh` -> `assess` -> `aggregate`), with prompt\u002Fimmutable dirs under\n  `configs\u002Feve\u002Fevaluation\u002Fcircle_packing\u002F`.\n",{"name":5,"description":7},"main","979065fd2cf1b3be9caf70596478fe27df7815ad","docs\u002Fskills\u002Fimplement-evaluation-steps",9,"https:\u002F\u002Fgithub.com\u002Fscaling-group\u002Feve","https:\u002F\u002Fgithub.com\u002Fscaling-group","https:\u002F\u002Fscaling-group.github.io\u002F","Department of Mathematics, National University of Singapore","2026-08-20 15:35:50",{"source":8,"skillName":5,"commands":41,"prompt":51,"zipAvailable":52},[42,45,48],{"tool":43,"command":44},"npx","npx skills add scaling-group\u002Feve --skill implement-evaluation-steps",{"tool":46,"command":47},"bunx","bunx skills add scaling-group\u002Feve --skill implement-evaluation-steps",{"tool":49,"command":50},"pnpm","pnpm dlx skills add scaling-group\u002Feve --skill implement-evaluation-steps","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fscaling-group\u002Feve\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Fimplement-evaluation-steps\u002FSKILL.md",true,[],[55,79,98,112,130,146],{"skillNo":56,"name":57,"nameZh":6,"description":58,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":59,"categories":60,"keywords":71,"publishedAt":26},"01M71RW4VG8A6FFRGT7H41J16J","writing-skills","Use when creating new skills, editing existing skills, or verifying skills work before deployment","https:\u002F\u002Fgithub.com\u002Fscaling-group\u002Feve\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Fwriting-skills\u002FSKILL.md",[61,66],{"categoryNo":62,"categoryName":63,"taxonomyType":64,"taxonomyTypeDesc":65},"01JBSEEDCAT000000000000018","写作与文案","SCENARIO","使用场景",{"categoryNo":67,"categoryName":68,"taxonomyType":69,"taxonomyTypeDesc":70},"01JBSEEDCAT000000000000036","内容创作者","PROFESSION","适用职业",[72,73,74,75,76,77,24,78,25],"writing","writing skills","work","editing","deployment","new","Existing",{"skillNo":80,"name":81,"nameZh":6,"description":82,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":83,"categories":84,"keywords":88,"publishedAt":26},"01M71RW2PFRVWH24AP25T6AV4E","theorem-search","Use TheoremSearch to find, verify, and cite mathematical theorem statements, named results, proof routes, and literature dependencies.","https:\u002F\u002Fgithub.com\u002Fscaling-group\u002Feve\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Ftheorem-search\u002FSKILL.md",[85],{"categoryNo":86,"categoryName":87,"taxonomyType":64,"taxonomyTypeDesc":65},"01JBSEEDCAT000000000000021","资料检索与研究",[89,90,24,91,92,93,94,95,96,97],"theorem search","search","literature","results","verify","find","dependencies","proof","Mathematical",{"skillNo":99,"name":100,"nameZh":6,"description":101,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":102,"categories":103,"keywords":104,"publishedAt":26},"01M71RW0N0GQR9GW1B3HKBB800","supervise-run","Use when monitoring a running Eve experiment and recovering from problems while it runs","https:\u002F\u002Fgithub.com\u002Fscaling-group\u002Feve\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Fsupervise-run\u002FSKILL.md",[],[105,106,107,108,109,110,111,25,24,23],"supervise run","supervise","run","monitoring","experiment","it","running",{"skillNo":113,"name":114,"nameZh":6,"description":115,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":116,"categories":117,"keywords":118,"publishedAt":26},"01M71RVYA101VVX3CKR2HZHBT3","run-math-proof-quickstart","Use when running a Math Proof quickstart from a user-provided problem: create a local task copy, launch a real EvE attempt, supervise it, and report proof\u002Fevaluation results.","https:\u002F\u002Fgithub.com\u002Fscaling-group\u002Feve\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Frun-math-proof-quickstart\u002FSKILL.md",[],[107,119,120,121,122,96,123,124,111,125,126,23,24,25,127,128,92,106,129,20],"math","run math proof quickstart","quickstart","math proof","copy","launch","problem","local","task","create","report",{"skillNo":131,"name":132,"nameZh":6,"description":133,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":134,"categories":135,"keywords":136,"publishedAt":26},"01M71RVWYKH6TF0WB6X57GMWJ2","run-circle-packing-smoke","Use when a fast end-to-end smoke run is needed for the Eve loop development, will run on circle packing task.","https:\u002F\u002Fgithub.com\u002Fscaling-group\u002Feve\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Frun-circle-packing-smoke\u002FSKILL.md",[],[137,138,107,139,140,25,141,142,143,23,24,144,127,145],"circle","packing","smoke","run circle packing smoke","development","for","end","loop","fast",{"skillNo":147,"name":148,"nameZh":6,"description":149,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":150,"categories":151,"keywords":155,"publishedAt":26},"01M71RVVCE1NEF4ZM405WX815Z","resume-run","Use when continuing an interrupted or paused Eve run from where it stopped (same run, same id), optionally after editing the experiment","https:\u002F\u002Fgithub.com\u002Fscaling-group\u002Feve\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Fresume-run\u002FSKILL.md",[152],{"categoryNo":153,"categoryName":154,"taxonomyType":69,"taxonomyTypeDesc":70},"01JBSEEDCAT000000000000038","人力资源",[156,107,157,75,23,24,25,158,159,109,110],"resume","resume run","id","same"]