[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M71RVVCE1NEF4ZM405WX815Z":3,"skill-same:01M71RVVCE1NEF4ZM405WX815Z":63},{"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":23,"publishedAt":35,"linkOnly":36,"markdownContent":37,"markdownBody":38,"frontMatter":39,"refName":40,"commitSha":41,"skillDirectory":42,"forksCount":43,"repositoryHtmlUrl":44,"repositoryHomepageUrl":6,"ownerHtmlUrl":45,"ownerBlogUrl":46,"ownerDescription":47,"updatedAt":48,"install":49,"resources":62},"01M71RVVCE1NEF4ZM405WX815Z","resume-run",null,"Use when continuing an interrupted or paused Eve run from where it stopped (same run, same id), optionally after editing the experiment","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\u002Fresume-run\u002FSKILL.md",[18],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000038","人力资源","PROFESSION","适用职业",[24,25,26,27,28,29,30,31,32,33,34],"resume run","resume","run","editing","eve","use","when","id","same","experiment","it","2026-11-10 13:02:02",false,"---\nname: resume-run\ndescription: \"Use when continuing an interrupted or paused Eve run from where it stopped (same run, same id), optionally after editing the experiment\"\n---\n\n# Resume a run\n\n`resume` continues the **same** run (same run id, same lineage) from the last completed iteration. To instead start a **new** run seeded from one or more prior runs, see `..\u002Fimport-run\u002FSKILL.md`.\n\nrun_root layout and the checkpoint\u002Fsnapshot model are described in `docs\u002Fusing-eve.md`.\n\n## 1. Pause (if the run is live)\n\nKill the process at any time — `Ctrl-C`, `SIGTERM`, or `tmux kill-session`. **Do not wait for a \"safe\" boundary.** Each completed iteration is snapshotted, so an interrupted iteration is recoverable; the cleanup happens on resume, not on kill.\n\n## 2. Adjust (optional, before resuming)\n\nThese take effect on resume (the run re-reads them from disk each iteration):\n\n- the optimizer's `immutable\u002F` and `prompt\u002F` files\n- the run config\n\n**Does NOT take effect on resume:** editing `initial_guidance\u002F`. The optimizer population is restored from the run's lineage DB, not re-seeded. To change the optimizer itself, start a fresh run instead — see `..\u002Fimport-run\u002FSKILL.md`.\n\n## 3. Resume\n\n```bash\nuv run python -m scaling_evolve.algorithms.eve.runner --config-name=\u003Cconfig> resume_from=\u003Cabsolute_run_root>\n```\n\nThe runner does the recovery for you: it reads `\u003Crun_root>\u002Fcheckpoint.json` for the last completed iteration, rolls the solver\u002Foptimizer lineage DBs back to that iteration's snapshot, restores local CSV telemetry to that same iteration, archives interrupted post-checkpoint workspaces under `\u003Crun_root>\u002F.resume_archive\u002F`, and continues under the **same run id**.\n\n- `resume_from` must be an **absolute** path; a relative path is rejected.\n- If you launched without wandb, re-pass `logger.wandb.enabled=false` here too (logger overrides are not persisted across the resume invocation).\n- Optional: `resume_iteration=\u003CK>` rolls back to an earlier completed iteration `K` (must be `\u003C= last_completed_iteration`).\n- Local telemetry under `\u003Crun_root>\u002Ftelemetry\u002F` is logical experiment output. After resume, it should contain rows through the checkpoint anchor plus newly completed post-resume rows, not rows from the failed partial attempt.\n- `.resume_archive\u002Fresume_*\u002Fmanifest.json` records any active post-anchor workspace paths that were moved out of the logical run surface. Use it when debugging the failed interrupted attempt.\n\n## 4. Verify (quick sanity)\n\n- An edit you made in step 2 shows up in the **post-resume** worker session logs\u002Ftranscripts (the optimize session under the new workspace), and is **absent** from already-completed ones. Prompt edits are reflected in the worker session, not copied verbatim into the workspace as a file.\n- The run continues under the original run id. Use `checkpoint.json`, lineage DBs, and `telemetry\u002F` for logical state; `runner.log` is an operational trace and may include failed pre-resume attempt output.\n- To read the resulting population, see `..\u002Finspect-population\u002FSKILL.md`.\n\n## Gotchas\n\n- Resume needs snapshots. They are written every iteration when `loop.enable_resume=true` (the default). A run launched with `loop.enable_resume=false` has no snapshots and **cannot** be resumed.\n- Do not delete `\u003Crun_root>\u002F.snapshots\u002F` or `checkpoint.json` — a missing snapshot makes resume fail hard.\n- `resume_from` and `import_from` are mutually exclusive; setting both is an error.\n- Remote wandb state is outside the local rollback contract. If wandb is enabled, treat it as remote operational logging, not as the source of truth for same-run resume.\n","# Resume a run\n\n`resume` continues the **same** run (same run id, same lineage) from the last completed iteration. To instead start a **new** run seeded from one or more prior runs, see `..\u002Fimport-run\u002FSKILL.md`.\n\nrun_root layout and the checkpoint\u002Fsnapshot model are described in `docs\u002Fusing-eve.md`.\n\n## 1. Pause (if the run is live)\n\nKill the process at any time — `Ctrl-C`, `SIGTERM`, or `tmux kill-session`. **Do not wait for a \"safe\" boundary.** Each completed iteration is snapshotted, so an interrupted iteration is recoverable; the cleanup happens on resume, not on kill.\n\n## 2. Adjust (optional, before resuming)\n\nThese take effect on resume (the run re-reads them from disk each iteration):\n\n- the optimizer's `immutable\u002F` and `prompt\u002F` files\n- the run config\n\n**Does NOT take effect on resume:** editing `initial_guidance\u002F`. The optimizer population is restored from the run's lineage DB, not re-seeded. To change the optimizer itself, start a fresh run instead — see `..\u002Fimport-run\u002FSKILL.md`.\n\n## 3. Resume\n\n```bash\nuv run python -m scaling_evolve.algorithms.eve.runner --config-name=\u003Cconfig> resume_from=\u003Cabsolute_run_root>\n```\n\nThe runner does the recovery for you: it reads `\u003Crun_root>\u002Fcheckpoint.json` for the last completed iteration, rolls the solver\u002Foptimizer lineage DBs back to that iteration's snapshot, restores local CSV telemetry to that same iteration, archives interrupted post-checkpoint workspaces under `\u003Crun_root>\u002F.resume_archive\u002F`, and continues under the **same run id**.\n\n- `resume_from` must be an **absolute** path; a relative path is rejected.\n- If you launched without wandb, re-pass `logger.wandb.enabled=false` here too (logger overrides are not persisted across the resume invocation).\n- Optional: `resume_iteration=\u003CK>` rolls back to an earlier completed iteration `K` (must be `\u003C= last_completed_iteration`).\n- Local telemetry under `\u003Crun_root>\u002Ftelemetry\u002F` is logical experiment output. After resume, it should contain rows through the checkpoint anchor plus newly completed post-resume rows, not rows from the failed partial attempt.\n- `.resume_archive\u002Fresume_*\u002Fmanifest.json` records any active post-anchor workspace paths that were moved out of the logical run surface. Use it when debugging the failed interrupted attempt.\n\n## 4. Verify (quick sanity)\n\n- An edit you made in step 2 shows up in the **post-resume** worker session logs\u002Ftranscripts (the optimize session under the new workspace), and is **absent** from already-completed ones. Prompt edits are reflected in the worker session, not copied verbatim into the workspace as a file.\n- The run continues under the original run id. Use `checkpoint.json`, lineage DBs, and `telemetry\u002F` for logical state; `runner.log` is an operational trace and may include failed pre-resume attempt output.\n- To read the resulting population, see `..\u002Finspect-population\u002FSKILL.md`.\n\n## Gotchas\n\n- Resume needs snapshots. They are written every iteration when `loop.enable_resume=true` (the default). A run launched with `loop.enable_resume=false` has no snapshots and **cannot** be resumed.\n- Do not delete `\u003Crun_root>\u002F.snapshots\u002F` or `checkpoint.json` — a missing snapshot makes resume fail hard.\n- `resume_from` and `import_from` are mutually exclusive; setting both is an error.\n- Remote wandb state is outside the local rollback contract. If wandb is enabled, treat it as remote operational logging, not as the source of truth for same-run resume.\n",{"name":5,"description":7},"main","979065fd2cf1b3be9caf70596478fe27df7815ad","docs\u002Fskills\u002Fresume-run",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":50,"prompt":60,"zipAvailable":61},[51,54,57],{"tool":52,"command":53},"npx","npx skills add scaling-group\u002Feve --skill resume-run",{"tool":55,"command":56},"bunx","bunx skills add scaling-group\u002Feve --skill resume-run",{"tool":58,"command":59},"pnpm","pnpm dlx skills add scaling-group\u002Feve --skill resume-run","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fscaling-group\u002Feve\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Fresume-run\u002FSKILL.md",true,[],[64,85,104,115,134,150],{"skillNo":65,"name":66,"nameZh":6,"description":67,"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":78,"publishedAt":35},"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",[70,75],{"categoryNo":71,"categoryName":72,"taxonomyType":73,"taxonomyTypeDesc":74},"01JBSEEDCAT000000000000018","写作与文案","SCENARIO","使用场景",{"categoryNo":76,"categoryName":77,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000036","内容创作者",[79,80,81,27,82,83,29,84,30],"writing","writing skills","work","deployment","new","Existing",{"skillNo":86,"name":87,"nameZh":6,"description":88,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":89,"categories":90,"keywords":94,"publishedAt":35},"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",[91],{"categoryNo":92,"categoryName":93,"taxonomyType":73,"taxonomyTypeDesc":74},"01JBSEEDCAT000000000000021","资料检索与研究",[95,96,29,97,98,99,100,101,102,103],"search","theorem search","literature","results","verify","find","dependencies","proof","Mathematical",{"skillNo":105,"name":106,"nameZh":6,"description":107,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":108,"categories":109,"keywords":110,"publishedAt":35},"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",[],[26,111,112,29,28,113,30,114,33,34],"supervise","supervise run","running","monitoring",{"skillNo":116,"name":117,"nameZh":6,"description":118,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":119,"categories":120,"keywords":121,"publishedAt":35},"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",[],[122,123,124,125,26,102,126,127,128,129,113,130,131,28,29,30,132,133,98,111],"math proof","quickstart","run math proof quickstart","math","Evaluation","copy","launch","report","problem","local","task","create",{"skillNo":135,"name":136,"nameZh":6,"description":137,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":138,"categories":139,"keywords":140,"publishedAt":35},"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",[],[141,142,143,26,144,28,145,146,147,29,148,30,132,149],"smoke","circle","packing","run circle packing smoke","development","for","end","loop","fast",{"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":156,"publishedAt":35},"01M71RVSM279Y6VCQV0ZWWTDR0","matlas-search","Use Matlas semantic search over published mathematical literature to find canonical statements, hypotheses, and citation metadata for standard or named results.","https:\u002F\u002Fgithub.com\u002Fscaling-group\u002Feve\u002Fblob\u002Fmain\u002Fdocs\u002Fskills\u002Fmatlas-search\u002FSKILL.md",[],[95,157,146,29,158,97,98,100,159,160,103,161,162,163],"matlas search","standard","semantic search","canonical","semantic","metadata","Citation"]