[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M790ZP5982T35TQCRETGDMBM":3,"skill-same:01M790ZP5982T35TQCRETGDMBM":66},{"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":26,"commitSha":27,"skillDirectory":28,"forksCount":29,"repositoryHtmlUrl":30,"repositoryHomepageUrl":6,"ownerHtmlUrl":31,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":32,"install":33,"resources":45},"01M790ZP5982T35TQCRETGDMBM","training-and-cli",null,"Runs and troubleshoots MuZero General CLI, Python API, Ray training\u002Ftesting, TensorBoard logging, and safe smoke workflows.","VectorSpaceLab\u002FAREX-Skill","VectorSpaceLab","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F180637102?v=4","01M7538C5WPSJGWF7KHR4EFZ9F","01M7538C6R2GFP53NPXTCZNBRT",351,"MIT","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fmuzero-general\u002Fsub-skills\u002Ftraining-and-cli\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: training-and-cli\ndescription: \"Runs and troubleshoots MuZero General CLI, Python API, Ray\n  training\u002Ftesting, TensorBoard logging, and safe smoke workflows.\"\ndisable-model-invocation: true\nmetadata:\n  disco-role: operating\nlicense: MIT\n---\n\n# Training and CLI\n\nUse this sub-skill when the task is to run MuZero General, build a safe training or testing command, use the `MuZero` Python API, adjust a JSON config override, understand Ray workers, inspect the upstream CartPole CI recipe, or debug training\u002Fruntime startup failures.\n\n## Fast route\n\n1. Read [training-workflows.md](references\u002Ftraining-workflows.md) for self-contained bundled-source entry points, Python API patterns, TensorBoard usage, hyperparameter search, and safe smoke overrides.\n2. Read [api-reference.md](references\u002Fapi-reference.md) when you need verified signatures, constructor behavior, worker roles, checkpoint lifecycle, or replay\u002Ftrainer\u002Fshared-storage interactions.\n3. Read [troubleshooting.md](references\u002Ftroubleshooting.md) for invalid game\u002Fconfig names, Ray startup failures, GPU resource conflicts, long-running training, TensorBoard confusion, Gym\u002FRay warnings, or interactive menu blocking.\n4. Run the bundled safe smoke helper before real training:\n\n   ```bash\n   python sub-skills\u002Ftraining-and-cli\u002Fscripts\u002Fmuzero_cli_smoke.py --game tictactoe --training-steps 0 --json\n   ```\n\n   The helper imports the bundled `runtime\u002Fsource\u002F` snapshot by default, constructs `MuZero`, applies CPU-safe overrides, reports key config\u002Fcheckpoint facts, and shuts Ray down. It does not train unless `--run-train` is explicitly supplied. Use `--repo-root` only for a staged editable copy or another target checkout.\n\n## Primary workflows\n\n- **Self-contained entry point:** from the skill root, use `python scripts\u002Frun_muzero.py --game \u003Cgame> --config-json '\u003Cjson-config>' --mode construct|train|test`. It runs against bundled `runtime\u002Fsource\u002F` unless `--repo-root` points to an intentionally staged copy or target checkout. Use JSON overrides for bounded runs such as `{\"training_steps\": 0, \"num_simulations\": 1, \"max_moves\": 1, \"save_model\": false}`.\n- **Python API run:** instantiate `MuZero(game_name, config=None, split_resources_in=1)`, optionally call `load_model(...)`, then call `train(log_in_tensorboard=True|False)` or `test(render=False, ...)`.\n- **Smoke before training:** prefer the bundled helper or `python scripts\u002Frun_muzero.py --game tictactoe --safe-smoke --mode construct --json` before launching Ray workers that can run indefinitely.\n- **Real training:** choose a game\u002Fconfig in sibling [games-and-configs](..\u002Fgames-and-configs\u002FSKILL.md), validate model\u002Fsearch shapes in [models-and-mcts](..\u002Fmodels-and-mcts\u002FSKILL.md), then run training with a deliberate result path and TensorBoard plan.\n- **Checkpoint-aware runs:** read [checkpoints-and-diagnostics](..\u002Fcheckpoints-and-diagnostics\u002FSKILL.md) before loading `model.checkpoint` or `replay_buffer.pkl` artifacts.\n\n## Safety boundaries\n\n- Do not run the upstream interactive menu (`runtime\u002Fsource\u002Fmuzero.py` with no args) in automation unless the user explicitly wants prompts. Prefer `scripts\u002Frun_muzero.py`, which has no interactive menu mode.\n- Do not run the upstream 7,500-step CartPole CI recipe as a normal smoke; it is an expensive native candidate, not the default verification path.\n- Do not enable `train_on_gpu`, `selfplay_on_gpu`, or `reanalyse_on_gpu` while setting `max_num_gpus = 0`.\n- Do not assume all optional games are installed. Atari, LunarLander\u002FBox2D, MiniGrid, and OpenSpiel need extra packages; route those checks to [games-and-configs](..\u002Fgames-and-configs\u002FSKILL.md).\n\n## When to stop and ask\n\nAsk for user approval before long training, installing optional game dependencies, downloading Atari ROM\u002Fassets, using GPUs for a heavy run, or overwriting\u002Fcheckpointing into a user-important results directory. Otherwise use bounded CPU-safe smoke checks first.\n","# Training and CLI\n\nUse this sub-skill when the task is to run MuZero General, build a safe training or testing command, use the `MuZero` Python API, adjust a JSON config override, understand Ray workers, inspect the upstream CartPole CI recipe, or debug training\u002Fruntime startup failures.\n\n## Fast route\n\n1. Read [training-workflows.md](references\u002Ftraining-workflows.md) for self-contained bundled-source entry points, Python API patterns, TensorBoard usage, hyperparameter search, and safe smoke overrides.\n2. Read [api-reference.md](references\u002Fapi-reference.md) when you need verified signatures, constructor behavior, worker roles, checkpoint lifecycle, or replay\u002Ftrainer\u002Fshared-storage interactions.\n3. Read [troubleshooting.md](references\u002Ftroubleshooting.md) for invalid game\u002Fconfig names, Ray startup failures, GPU resource conflicts, long-running training, TensorBoard confusion, Gym\u002FRay warnings, or interactive menu blocking.\n4. Run the bundled safe smoke helper before real training:\n\n   ```bash\n   python sub-skills\u002Ftraining-and-cli\u002Fscripts\u002Fmuzero_cli_smoke.py --game tictactoe --training-steps 0 --json\n   ```\n\n   The helper imports the bundled `runtime\u002Fsource\u002F` snapshot by default, constructs `MuZero`, applies CPU-safe overrides, reports key config\u002Fcheckpoint facts, and shuts Ray down. It does not train unless `--run-train` is explicitly supplied. Use `--repo-root` only for a staged editable copy or another target checkout.\n\n## Primary workflows\n\n- **Self-contained entry point:** from the skill root, use `python scripts\u002Frun_muzero.py --game \u003Cgame> --config-json '\u003Cjson-config>' --mode construct|train|test`. It runs against bundled `runtime\u002Fsource\u002F` unless `--repo-root` points to an intentionally staged copy or target checkout. Use JSON overrides for bounded runs such as `{\"training_steps\": 0, \"num_simulations\": 1, \"max_moves\": 1, \"save_model\": false}`.\n- **Python API run:** instantiate `MuZero(game_name, config=None, split_resources_in=1)`, optionally call `load_model(...)`, then call `train(log_in_tensorboard=True|False)` or `test(render=False, ...)`.\n- **Smoke before training:** prefer the bundled helper or `python scripts\u002Frun_muzero.py --game tictactoe --safe-smoke --mode construct --json` before launching Ray workers that can run indefinitely.\n- **Real training:** choose a game\u002Fconfig in sibling [games-and-configs](..\u002Fgames-and-configs\u002FSKILL.md), validate model\u002Fsearch shapes in [models-and-mcts](..\u002Fmodels-and-mcts\u002FSKILL.md), then run training with a deliberate result path and TensorBoard plan.\n- **Checkpoint-aware runs:** read [checkpoints-and-diagnostics](..\u002Fcheckpoints-and-diagnostics\u002FSKILL.md) before loading `model.checkpoint` or `replay_buffer.pkl` artifacts.\n\n## Safety boundaries\n\n- Do not run the upstream interactive menu (`runtime\u002Fsource\u002Fmuzero.py` with no args) in automation unless the user explicitly wants prompts. Prefer `scripts\u002Frun_muzero.py`, which has no interactive menu mode.\n- Do not run the upstream 7,500-step CartPole CI recipe as a normal smoke; it is an expensive native candidate, not the default verification path.\n- Do not enable `train_on_gpu`, `selfplay_on_gpu`, or `reanalyse_on_gpu` while setting `max_num_gpus = 0`.\n- Do not assume all optional games are installed. Atari, LunarLander\u002FBox2D, MiniGrid, and OpenSpiel need extra packages; route those checks to [games-and-configs](..\u002Fgames-and-configs\u002FSKILL.md).\n\n## When to stop and ask\n\nAsk for user approval before long training, installing optional game dependencies, downloading Atari ROM\u002Fassets, using GPUs for a heavy run, or overwriting\u002Fcheckpointing into a user-important results directory. Otherwise use bounded CPU-safe smoke checks first.\n",{"name":5,"license":14,"metadata":23,"description":7,"disable-model-invocation":25},{"disco-role":24},"operating",true,"main","ac3fe1afa80fb9a09775ecfb2b6cc3ba850a2db6","skills\u002Frepositories\u002Frepo-skills\u002Fmuzero-general\u002Fsub-skills\u002Ftraining-and-cli",27,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab","2026-09-03 10:25:21",{"source":8,"skillName":5,"commands":34,"prompt":44,"zipAvailable":25},[35,38,41],{"tool":36,"command":37},"npx","npx skills add VectorSpaceLab\u002FAREX-Skill --skill training-and-cli",{"tool":39,"command":40},"bunx","bunx skills add VectorSpaceLab\u002FAREX-Skill --skill training-and-cli",{"tool":42,"command":43},"pnpm","pnpm dlx skills add VectorSpaceLab\u002FAREX-Skill --skill training-and-cli","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fmuzero-general\u002Fsub-skills\u002Ftraining-and-cli\u002FSKILL.md",[46,52,56,60],{"path":47,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":50,"sourceHtmlUrl":51},"references\u002Fapi-reference.md","REFERENCE","参考文档",5178,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fmuzero-general\u002Fsub-skills\u002Ftraining-and-cli\u002Freferences\u002Fapi-reference.md",{"path":53,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":54,"sourceHtmlUrl":55},"references\u002Ftraining-workflows.md",7693,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fmuzero-general\u002Fsub-skills\u002Ftraining-and-cli\u002Freferences\u002Ftraining-workflows.md",{"path":57,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":58,"sourceHtmlUrl":59},"references\u002Ftroubleshooting.md",4600,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fmuzero-general\u002Fsub-skills\u002Ftraining-and-cli\u002Freferences\u002Ftroubleshooting.md",{"path":61,"resourceType":62,"resourceTypeDesc":63,"sizeBytes":64,"sourceHtmlUrl":65},"scripts\u002Fmuzero_cli_smoke.py","SCRIPT","脚本",7650,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fmuzero-general\u002Fsub-skills\u002Ftraining-and-cli\u002Fscripts\u002Fmuzero_cli_smoke.py",[67,75,82,89,96,103],{"skillNo":68,"name":69,"nameZh":6,"description":70,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":71,"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.","Apache-2.0","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":71,"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":71,"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":71,"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":71,"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":71,"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",[],[]]