[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M78SXM97FJMZEWJN8RW9F0M9":3,"skill-same:01M78SXM97FJMZEWJN8RW9F0M9":67},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":27,"commitSha":28,"skillDirectory":29,"forksCount":30,"repositoryHtmlUrl":31,"repositoryHomepageUrl":6,"ownerHtmlUrl":32,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":33,"install":34,"resources":46},"01M78SXM97FJMZEWJN8RW9F0M9","gpu-and-resource-operations",null,"Detect and select local NVIDIA GPUs, report their resource status, and explain the explicitly opt-in GPU keep-alive without hiding hardware or CUDA limitations.","VectorSpaceLab\u002FAREX-Skill","VectorSpaceLab","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F180637102?v=4","01M7538C5WPSJGWF7KHR4EFZ9F","01M7538C6R2GFP53NPXTCZNBRT",351,"Apache-2.0","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fauto-deep-researcher-24x7\u002Fsub-skills\u002Fgpu-and-resource-operations\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: gpu-and-resource-operations\ndescription: \"Detect and select local NVIDIA GPUs, report their resource status, and explain the explicitly opt-in GPU keep-alive without hiding hardware or CUDA limitations.\"\ndisable-model-invocation: true\nmetadata:\n  disco-role: operating\nlicense: Apache 2.0\n---\n\n# GPU and resource operations\n\nUse this skill for **local, read-only GPU inspection and selection** before an\nexperiment. It covers NVIDIA visibility, memory-based availability, the\n`reserve_last` policy, and the optional keep-alive's safety contract. It does\nnot launch training, reserve a device through a scheduler, or start a keeper\nimplicitly.\n\n## Safety contract\n\n- Begin with the bundled `scripts\u002Fgpu_status.py`; it only invokes\n  `nvidia-smi` queries and never allocates CUDA memory.\n- Treat an absent, timed-out, or failed `nvidia-smi` query as **no usable GPU**.\n  Never infer free devices from configuration, PyTorch importability, or a\n  missing status result.\n- “Free” means reported `memory_used_mb \u003C 1000` by default, not that a device is\n  exclusively reserved. Confirm ownership before launching work.\n- `reserve_last=true` is a selection policy for a keep-alive slot, not a lock.\n  With one detected GPU it remains usable; with two or more, the highest listed\n  index is excluded from experiment candidates.\n- Do not run the keep-alive during inspection. It allocates a CUDA tensor and\n  performs repeated operations, so it is an explicit side effect requiring\n  user approval.\n\n## Operating flow\n\n1. Run `python scripts\u002Fgpu_status.py --help` if the interface is unfamiliar;\n   help must not probe hardware.\n2. Run `python scripts\u002Fgpu_status.py` (or add `--json` for machine-readable\n   output). Record the status reason when no GPU is reported.\n3. Select only from `usable_gpus` and then filter to `free_gpus`. Keep the\n   reserved device out of training when the default policy excludes it.\n4. If the user explicitly chooses a device, report its `gpu_id`, name, memory,\n   utilization, and temperature before passing its ID to the experiment\n   launcher. Local launchers express this as `CUDA_VISIBLE_DEVICES`.\n5. Before CUDA work, verify both a working NVIDIA driver (`nvidia-smi`) and a\n   PyTorch build for which `torch.cuda.is_available()` is true. A Python\n   package alone is not proof of CUDA support.\n6. For SSH\u002FSlurm status, scheduler allocation, or transport details, route to\n   `execution-and-monitoring`. For loop launch and experiment lifecycle, route\n   to `autonomous-experiments`. For installing external skills, route to\n   `skills-and-installation`.\n\n## Bundled references\n\n- [GPU status and selection](references\u002Fgpu-status.md) — exact detector\n  signatures, output shapes, query fields, thresholds, and edge cases.\n- [Keep-alive contract](references\u002Fkeeper.md) — explicit side effects,\n  PyTorch\u002FCUDA prerequisites, interval, signals, cleanup, and invocation\n  rules.\n- [Troubleshooting](references\u002Ftroubleshooting.md) — predictable driver,\n  CUDA, parsing, selection, and keeper failures.\n\n## Output contract\n\nA normal report should include:\n\n- `detected`: whether a complete local NVIDIA status was obtained;\n- one row per GPU with ID, name, used\u002Ftotal memory in MB, utilization percent,\n  and temperature in °C;\n- `usable_gpus` after `reserve_last` policy;\n- `free_gpus` after the strict 1000-MB default threshold (or the stated\n  override);\n- an explicit `reason` when detection\u002Fstatus is unavailable; and\n- whether a last device was excluded, without claiming a reservation lock.\n\nWhen `detected` is false, both usable and free lists must be empty. Do not\nsilently substitute Slurm queue occupancy or a remote host's status for local\nNVIDIA measurements.\n\nIf a report is used to choose a launch mask, preserve the host GPU IDs and the\nselection policy in the handoff. Re-run the read-only probe immediately before\na long launch because memory and utilization are live observations, not a\nreservation. Never claim that a keeper is active unless its explicit process\nand stop\u002Fcleanup outcome have been checked.\n","# GPU and resource operations\n\nUse this skill for **local, read-only GPU inspection and selection** before an\nexperiment. It covers NVIDIA visibility, memory-based availability, the\n`reserve_last` policy, and the optional keep-alive's safety contract. It does\nnot launch training, reserve a device through a scheduler, or start a keeper\nimplicitly.\n\n## Safety contract\n\n- Begin with the bundled `scripts\u002Fgpu_status.py`; it only invokes\n  `nvidia-smi` queries and never allocates CUDA memory.\n- Treat an absent, timed-out, or failed `nvidia-smi` query as **no usable GPU**.\n  Never infer free devices from configuration, PyTorch importability, or a\n  missing status result.\n- “Free” means reported `memory_used_mb \u003C 1000` by default, not that a device is\n  exclusively reserved. Confirm ownership before launching work.\n- `reserve_last=true` is a selection policy for a keep-alive slot, not a lock.\n  With one detected GPU it remains usable; with two or more, the highest listed\n  index is excluded from experiment candidates.\n- Do not run the keep-alive during inspection. It allocates a CUDA tensor and\n  performs repeated operations, so it is an explicit side effect requiring\n  user approval.\n\n## Operating flow\n\n1. Run `python scripts\u002Fgpu_status.py --help` if the interface is unfamiliar;\n   help must not probe hardware.\n2. Run `python scripts\u002Fgpu_status.py` (or add `--json` for machine-readable\n   output). Record the status reason when no GPU is reported.\n3. Select only from `usable_gpus` and then filter to `free_gpus`. Keep the\n   reserved device out of training when the default policy excludes it.\n4. If the user explicitly chooses a device, report its `gpu_id`, name, memory,\n   utilization, and temperature before passing its ID to the experiment\n   launcher. Local launchers express this as `CUDA_VISIBLE_DEVICES`.\n5. Before CUDA work, verify both a working NVIDIA driver (`nvidia-smi`) and a\n   PyTorch build for which `torch.cuda.is_available()` is true. A Python\n   package alone is not proof of CUDA support.\n6. For SSH\u002FSlurm status, scheduler allocation, or transport details, route to\n   `execution-and-monitoring`. For loop launch and experiment lifecycle, route\n   to `autonomous-experiments`. For installing external skills, route to\n   `skills-and-installation`.\n\n## Bundled references\n\n- [GPU status and selection](references\u002Fgpu-status.md) — exact detector\n  signatures, output shapes, query fields, thresholds, and edge cases.\n- [Keep-alive contract](references\u002Fkeeper.md) — explicit side effects,\n  PyTorch\u002FCUDA prerequisites, interval, signals, cleanup, and invocation\n  rules.\n- [Troubleshooting](references\u002Ftroubleshooting.md) — predictable driver,\n  CUDA, parsing, selection, and keeper failures.\n\n## Output contract\n\nA normal report should include:\n\n- `detected`: whether a complete local NVIDIA status was obtained;\n- one row per GPU with ID, name, used\u002Ftotal memory in MB, utilization percent,\n  and temperature in °C;\n- `usable_gpus` after `reserve_last` policy;\n- `free_gpus` after the strict 1000-MB default threshold (or the stated\n  override);\n- an explicit `reason` when detection\u002Fstatus is unavailable; and\n- whether a last device was excluded, without claiming a reservation lock.\n\nWhen `detected` is false, both usable and free lists must be empty. Do not\nsilently substitute Slurm queue occupancy or a remote host's status for local\nNVIDIA measurements.\n\nIf a report is used to choose a launch mask, preserve the host GPU IDs and the\nselection policy in the handoff. Re-run the read-only probe immediately before\na long launch because memory and utilization are live observations, not a\nreservation. Never claim that a keeper is active unless its explicit process\nand stop\u002Fcleanup outcome have been checked.\n",{"name":5,"license":23,"metadata":24,"description":7,"disable-model-invocation":26},"Apache 2.0",{"disco-role":25},"operating",true,"main","ac3fe1afa80fb9a09775ecfb2b6cc3ba850a2db6","skills\u002Frepositories\u002Frepo-skills\u002Fauto-deep-researcher-24x7\u002Fsub-skills\u002Fgpu-and-resource-operations",27,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab","2026-09-03 10:25:21",{"source":8,"skillName":5,"commands":35,"prompt":45,"zipAvailable":26},[36,39,42],{"tool":37,"command":38},"npx","npx skills add VectorSpaceLab\u002FAREX-Skill --skill gpu-and-resource-operations",{"tool":40,"command":41},"bunx","bunx skills add VectorSpaceLab\u002FAREX-Skill --skill gpu-and-resource-operations",{"tool":43,"command":44},"pnpm","pnpm dlx skills add VectorSpaceLab\u002FAREX-Skill --skill gpu-and-resource-operations","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fauto-deep-researcher-24x7\u002Fsub-skills\u002Fgpu-and-resource-operations\u002FSKILL.md",[47,53,57,61],{"path":48,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":51,"sourceHtmlUrl":52},"references\u002Fgpu-status.md","REFERENCE","参考文档",4232,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fauto-deep-researcher-24x7\u002Fsub-skills\u002Fgpu-and-resource-operations\u002Freferences\u002Fgpu-status.md",{"path":54,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":55,"sourceHtmlUrl":56},"references\u002Fkeeper.md",3848,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fauto-deep-researcher-24x7\u002Fsub-skills\u002Fgpu-and-resource-operations\u002Freferences\u002Fkeeper.md",{"path":58,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":59,"sourceHtmlUrl":60},"references\u002Ftroubleshooting.md",3762,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fauto-deep-researcher-24x7\u002Fsub-skills\u002Fgpu-and-resource-operations\u002Freferences\u002Ftroubleshooting.md",{"path":62,"resourceType":63,"resourceTypeDesc":64,"sizeBytes":65,"sourceHtmlUrl":66},"scripts\u002Fgpu_status.py","SCRIPT","脚本",8082,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fauto-deep-researcher-24x7\u002Fsub-skills\u002Fgpu-and-resource-operations\u002Fscripts\u002Fgpu_status.py",[68,75,82,89,96,103],{"skillNo":69,"name":70,"nameZh":6,"description":71,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":72,"categories":73,"keywords":74,"publishedAt":18},"01M799NYSH1R3RVAHSM2D88Y84","passnet-triton-opt","Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block\u002Fgrid\u002Fwarp tuning, autotune policy, launch overhead, and per-op numeric recipes to pass the dtype baseline tolerances. Use when a pass matches and is correct but speedup below expected, or when correctness fails by small numeric margins.","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-triton-opt\u002FSKILL.md",[],[],{"skillNo":76,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":18},"01M799NS5APEC0N1FZ7XZVW6BN","passnet-solve","END-TO-END playbook for solving one PassNet sample: analyze the computation graphs, decide the optimization strategy (what to fuse, what to replace, what to leave alone), drive the iteration loop, and maximize the sample score. This is the ENTRY skill — invoke it first for any PassNet optimization task; it tells you when to use passnet-pattern-fusion, passnet-triton-opt and passnet-feedback.","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-solve\u002FSKILL.md",[],[],{"skillNo":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":18},"01M799NNY9BKSM21V5R7QVTBJX","passnet-skill","PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under .\u002Fpass_dir\u002F, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-skill\u002FSKILL.md",[],[],{"skillNo":90,"name":91,"nameZh":6,"description":92,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":93,"categories":94,"keywords":95,"publishedAt":18},"01M799NHH6NP7Z1RK77KGXAN6S","passnet-pattern-fusion","Author PassNet pass files: write patterns that actually MATCH the FX graph, pick fusion regions (which ops to absorb into one Triton kernel), and structure multi-pass submissions with the mandatory shared-dispatch architecture. Use when creating or fixing pattern()\u002Freplacement_args()\u002Freplacement_func() files, when a pass \"failed to match\", or when deciding how to fuse multiple kernels\u002Fops.","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-pattern-fusion\u002FSKILL.md",[],[],{"skillNo":97,"name":98,"nameZh":6,"description":99,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":100,"categories":101,"keywords":102,"publishedAt":18},"01M799ND2J4WNC2PGAZ0A06NDF","passnet-orchestrate","Round\u002Fbatch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep\u002Frevert decisions. For solving one concrete sample end to end, use passnet-solve as the entry skill; consult this skill only when you need broader planning or a specific decision gate.","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-orchestrate\u002FSKILL.md",[],[],{"skillNo":104,"name":105,"nameZh":6,"description":106,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":107,"categories":108,"keywords":109,"publishedAt":18},"01M799N8SBTMY02ZDG2AZF5ZWE","passnet-feedback","Fast, deterministic feedback for PassNet work: pre-flight pattern\u002Fmatch verification WITHOUT burning a GPU evaluation, per-node bottleneck analysis of a sample's graphs, and eval-log parsing into per-variant status + estimated score + failure classification. Use BEFORE every GPU evaluation (check_pattern), at the START of a sample (analyze_graph), and AFTER every evaluation (parse_eval_log).","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-feedback\u002FSKILL.md",[],[]]