[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M79119T4TYEEPAAV0SK0STR3":3,"skill-same:01M79119T4TYEEPAAV0SK0STR3":63},{"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},"01M79119T4TYEEPAAV0SK0STR3","simulation-rendering",null,"Use for MuJoCo model loading, simulation state and time inspection, headless or onscreen rendering, camera\u002Foutput configuration, and viewer\u002Fdisplay diagnostics; route environment catalog, XML editing\u002FIK, MJX acceleration, and training elsewhere.","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\u002Fmyosuite\u002Fsub-skills\u002Fsimulation-rendering\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: \"simulation-rendering\"\ndescription: \"Use for MuJoCo model loading, simulation state and time\n  inspection, headless or onscreen rendering, camera\u002Foutput configuration, and\n  viewer\u002Fdisplay diagnostics; route environment catalog, XML editing\u002FIK, MJX\n  acceleration, and training elsewhere.\"\ndisable-model-invocation: true\nmetadata:\n  disco-role: operating\nlicense: Apache 2.0\n---\n\n# Simulation and rendering\n\nUse this sub-skill when the task concerns the MuJoCo model\u002Fdata handles behind a\nMyoSuite environment, rendering pixels or depth, selecting cameras, preserving\nor restoring simulator state, or diagnosing a viewer\u002Fdisplay failure. It covers\nthe base MuJoCo backend in MyoSuite 2.x. It does **not** teach task selection,\nXML mutation or IK, MJX\u002FJAX acceleration, or RL training.\n\n## Route first\n\n- Choose [environment workflows](..\u002Fenvironments\u002FSKILL.md) for registry lookup,\ntask configuration, reset\u002Fstep semantics, or environment-specific observation\nand reward behavior.\n- Choose [model editing and kinematics](..\u002Fmodel-editing-kinematics\u002FSKILL.md)\nfor XML changes, `MjSpec` editing, site\u002Fbody edits as a modeling operation, or\nIK.\n- Choose [MJX acceleration](..\u002Fmjx-acceleration\u002FSKILL.md) for JAX, MJX, CUDA,\nbatching, or accelerator performance.\n- Choose [training integration](..\u002Ftraining-integration\u002FSKILL.md) for policies,\nlearners, checkpoints, or long-running experiments.\n\n## Fast decision: which rendering path?\n\n1. **Headless or CI:** load `mujoco.MjModel`\u002F`mujoco.MjData`, step with\n   `mujoco.mj_step`, and use `MJRenderer.render_offscreen(...)`, or run the\n   bundled safe checker:\n   `python scripts\u002Fcheck_mujoco_xml.py --help`.\n2. **A window is explicitly wanted:** `env.mj_render()` delegates to\n   `env.mj_renderer.render_to_window()`, which creates a native passive viewer\n   on first use. This is display-dependent and is not verification ground truth.\n3. **Raw XML diagnosis:** run\n   `python scripts\u002Fcheck_mujoco_xml.py --xml MODEL.xml --render none` first.\n   It loads and steps without creating a viewer. Add `--render offscreen` only\n   when pixel output is required.\n\nNever use `myosuite.utils.examine_sim` as a headless check: its documented\nworkflow calls `mujoco.viewer.launch(...)` and is intentionally onscreen. Its\nsafe native candidate is `python -m myosuite.utils.examine_sim --help` only.\n\n## Environment rendering contract\n\n```python\nfrom myosuite.utils import gym\n\nenv = gym.make(\"myoElbowPose1D6MRandom-v0\")\nenv.reset(seed=1234)\nframe = env.mj_renderer.render_offscreen(\n    width=320, height=240, camera_id=-1, rgb=True\n)\nenv.close()\n```\n\n- `env.mj_model` is the compiled `mujoco.MjModel`; `env.mj_data` is its live\n  `mujoco.MjData`. The base environment constructs `env.mj_renderer` as an\n  `MJRenderer` for those handles.\n- `camera_id=-1` means MuJoCo's free camera. A named camera string or numeric\n  camera id selects a model camera. Use a model camera name only after checking\n  that it exists; a missing name is a model\u002Frendering error, not a display fix.\n- `render_offscreen` returns an RGB `numpy` array for the default call. The\n  detailed return matrix for RGB\u002Fdepth\u002Fsegmentation is in\n  [rendering-api.md](references\u002Frendering-api.md).\n- `env.mj_render()` is deliberately different: it opens\u002Fsynchronizes a window\n  and returns no frame. Do not call it in a server, CI job, or headless synthetic\n  case.\n- `env.viewer_setup(distance=..., azimuth=..., elevation=..., lookat=...,\n  render_actuator=..., render_tendon=...)` stores free-camera and visualization\n  settings for the next window or offscreen scene. Camera\u002Foutput details and\n  the distance adjustment are documented in [rendering-api.md](references\u002Frendering-api.md).\n\n## Deterministic headless rollout\n\nFor a raw model, the safe contract is:\n\n```bash\npython scripts\u002Fcheck_mujoco_xml.py \\\n  --xml MODEL.xml --render offscreen --frames 8 \\\n  --width 320 --height 240 --output-dir .\u002Fmujoco-check\n```\n\nThe command never calls `mujoco.viewer` or `launch_passive`. It prints model\nsizes, steps the requested number of frames, and writes one binary PPM image per\nrendered frame under the supplied output directory (for example,\n`mujoco-check\u002Fframe-0000.ppm`). `--render none` is the default and creates no\noutput directory. The script accepts optional comma-separated `--qpos` and\n`--ctrl`; it validates their lengths before applying them. See the script's\n`--help` output for the complete input\u002Foutput contract.\n\nFor a policy\u002Fenvironment rollout, use `MujocoEnv.examine_policy_new` only when\nits policy API is already available. Its supported `render` values are\n`\"onscreen\"`, `\"offscreen\"`, and `\"none\"`; offscreen mode accumulates RGB\nframes and writes an MP4 per episode through `imageio`. Set an explicit\n`output_dir`, `filename`, `frame_size`, and finite `horizon`. The safe raw-model\nscript is preferred for a small, dependency-light rendering check.\n\n## State and time workflow\n\n- `env.dt` is `env.mj_model.opt.timestep * env.frame_skip`.\n- `env.time` reads the observed simulation data time\n  (`env.obsd_mj_data.time`), so compare it with `env.mj_data.time` only when the\n  environment uses the same observed and ground-truth model.\n- `env.get_env_state()` returns copied `time`, `qpos`, `qvel`, optional `act`,\n  optional mocap position\u002Fquaternion, optional site position\u002Fquaternion, and\n  body position\u002Fquaternion arrays.\n- Save the returned dictionary before a branch, then call\n  `env.set_env_state(state)` to restore the base state. The method updates both\n  ground-truth and observed data where present and performs a MuJoCo step to\n  refresh derived state; treat restoration as a simulator-state operation, not\n  as a byte-for-byte snapshot of every internal buffer.\n- `env.reset(...)` establishes the task's initial state. After changing raw\n  `qpos`\u002F`qvel`, use `mujoco.mj_forward(model, data)` before inspecting derived\n  positions or rendering. Do not edit model arrays merely to move a body unless\n  the task is explicitly a model-editing task.\n\n## Visual observations\n\nVisual observations are opt-in. Call `env.get_obs(update_exteroception=True)`\nor `env.get_visuals(...)` when the environment has configured `visual_keys`.\nSupported key forms are `rgb:CAMERA:HxW:1d`, `rgb:CAMERA:HxW:2d`, and optional\nencoder forms such as `r3m18`, `r3m34`, `r3m50`, `rrl`, or `vc1` when their\noptional encoder dependencies are installed. A matching `d:` key requests\ndepth alongside the RGB key. `get_obs()` does not refresh exteroception by\ndefault, and `env_info[\"visual_dict\"]` may therefore be empty or stale unless\nvisuals were explicitly updated at the current simulation time.\n\n## Backend and safety boundary\n\nBase MuJoCo model loading, stepping, state inspection, and the `MJRenderer`\noffscreen path are CPU\u002Fbase-package capabilities, subject to a usable graphics\nbackend for pixel rendering. Onscreen viewing additionally requires a display\nand native viewer support. MJX\u002FJAX\u002FCUDA is optional and is not established by\nan offscreen CPU check; route it to the MJX sub-skill and report it separately.\n\nBefore reporting success, check the actual artifact: model load\u002Fstep succeeded,\nrendered arrays have the requested dimensions, and every requested output file\nexists and is non-empty. For failure symptoms and recovery branches, use\n[troubleshooting.md](references\u002Ftroubleshooting.md).","# Simulation and rendering\n\nUse this sub-skill when the task concerns the MuJoCo model\u002Fdata handles behind a\nMyoSuite environment, rendering pixels or depth, selecting cameras, preserving\nor restoring simulator state, or diagnosing a viewer\u002Fdisplay failure. It covers\nthe base MuJoCo backend in MyoSuite 2.x. It does **not** teach task selection,\nXML mutation or IK, MJX\u002FJAX acceleration, or RL training.\n\n## Route first\n\n- Choose [environment workflows](..\u002Fenvironments\u002FSKILL.md) for registry lookup,\ntask configuration, reset\u002Fstep semantics, or environment-specific observation\nand reward behavior.\n- Choose [model editing and kinematics](..\u002Fmodel-editing-kinematics\u002FSKILL.md)\nfor XML changes, `MjSpec` editing, site\u002Fbody edits as a modeling operation, or\nIK.\n- Choose [MJX acceleration](..\u002Fmjx-acceleration\u002FSKILL.md) for JAX, MJX, CUDA,\nbatching, or accelerator performance.\n- Choose [training integration](..\u002Ftraining-integration\u002FSKILL.md) for policies,\nlearners, checkpoints, or long-running experiments.\n\n## Fast decision: which rendering path?\n\n1. **Headless or CI:** load `mujoco.MjModel`\u002F`mujoco.MjData`, step with\n   `mujoco.mj_step`, and use `MJRenderer.render_offscreen(...)`, or run the\n   bundled safe checker:\n   `python scripts\u002Fcheck_mujoco_xml.py --help`.\n2. **A window is explicitly wanted:** `env.mj_render()` delegates to\n   `env.mj_renderer.render_to_window()`, which creates a native passive viewer\n   on first use. This is display-dependent and is not verification ground truth.\n3. **Raw XML diagnosis:** run\n   `python scripts\u002Fcheck_mujoco_xml.py --xml MODEL.xml --render none` first.\n   It loads and steps without creating a viewer. Add `--render offscreen` only\n   when pixel output is required.\n\nNever use `myosuite.utils.examine_sim` as a headless check: its documented\nworkflow calls `mujoco.viewer.launch(...)` and is intentionally onscreen. Its\nsafe native candidate is `python -m myosuite.utils.examine_sim --help` only.\n\n## Environment rendering contract\n\n```python\nfrom myosuite.utils import gym\n\nenv = gym.make(\"myoElbowPose1D6MRandom-v0\")\nenv.reset(seed=1234)\nframe = env.mj_renderer.render_offscreen(\n    width=320, height=240, camera_id=-1, rgb=True\n)\nenv.close()\n```\n\n- `env.mj_model` is the compiled `mujoco.MjModel`; `env.mj_data` is its live\n  `mujoco.MjData`. The base environment constructs `env.mj_renderer` as an\n  `MJRenderer` for those handles.\n- `camera_id=-1` means MuJoCo's free camera. A named camera string or numeric\n  camera id selects a model camera. Use a model camera name only after checking\n  that it exists; a missing name is a model\u002Frendering error, not a display fix.\n- `render_offscreen` returns an RGB `numpy` array for the default call. The\n  detailed return matrix for RGB\u002Fdepth\u002Fsegmentation is in\n  [rendering-api.md](references\u002Frendering-api.md).\n- `env.mj_render()` is deliberately different: it opens\u002Fsynchronizes a window\n  and returns no frame. Do not call it in a server, CI job, or headless synthetic\n  case.\n- `env.viewer_setup(distance=..., azimuth=..., elevation=..., lookat=...,\n  render_actuator=..., render_tendon=...)` stores free-camera and visualization\n  settings for the next window or offscreen scene. Camera\u002Foutput details and\n  the distance adjustment are documented in [rendering-api.md](references\u002Frendering-api.md).\n\n## Deterministic headless rollout\n\nFor a raw model, the safe contract is:\n\n```bash\npython scripts\u002Fcheck_mujoco_xml.py \\\n  --xml MODEL.xml --render offscreen --frames 8 \\\n  --width 320 --height 240 --output-dir .\u002Fmujoco-check\n```\n\nThe command never calls `mujoco.viewer` or `launch_passive`. It prints model\nsizes, steps the requested number of frames, and writes one binary PPM image per\nrendered frame under the supplied output directory (for example,\n`mujoco-check\u002Fframe-0000.ppm`). `--render none` is the default and creates no\noutput directory. The script accepts optional comma-separated `--qpos` and\n`--ctrl`; it validates their lengths before applying them. See the script's\n`--help` output for the complete input\u002Foutput contract.\n\nFor a policy\u002Fenvironment rollout, use `MujocoEnv.examine_policy_new` only when\nits policy API is already available. Its supported `render` values are\n`\"onscreen\"`, `\"offscreen\"`, and `\"none\"`; offscreen mode accumulates RGB\nframes and writes an MP4 per episode through `imageio`. Set an explicit\n`output_dir`, `filename`, `frame_size`, and finite `horizon`. The safe raw-model\nscript is preferred for a small, dependency-light rendering check.\n\n## State and time workflow\n\n- `env.dt` is `env.mj_model.opt.timestep * env.frame_skip`.\n- `env.time` reads the observed simulation data time\n  (`env.obsd_mj_data.time`), so compare it with `env.mj_data.time` only when the\n  environment uses the same observed and ground-truth model.\n- `env.get_env_state()` returns copied `time`, `qpos`, `qvel`, optional `act`,\n  optional mocap position\u002Fquaternion, optional site position\u002Fquaternion, and\n  body position\u002Fquaternion arrays.\n- Save the returned dictionary before a branch, then call\n  `env.set_env_state(state)` to restore the base state. The method updates both\n  ground-truth and observed data where present and performs a MuJoCo step to\n  refresh derived state; treat restoration as a simulator-state operation, not\n  as a byte-for-byte snapshot of every internal buffer.\n- `env.reset(...)` establishes the task's initial state. After changing raw\n  `qpos`\u002F`qvel`, use `mujoco.mj_forward(model, data)` before inspecting derived\n  positions or rendering. Do not edit model arrays merely to move a body unless\n  the task is explicitly a model-editing task.\n\n## Visual observations\n\nVisual observations are opt-in. Call `env.get_obs(update_exteroception=True)`\nor `env.get_visuals(...)` when the environment has configured `visual_keys`.\nSupported key forms are `rgb:CAMERA:HxW:1d`, `rgb:CAMERA:HxW:2d`, and optional\nencoder forms such as `r3m18`, `r3m34`, `r3m50`, `rrl`, or `vc1` when their\noptional encoder dependencies are installed. A matching `d:` key requests\ndepth alongside the RGB key. `get_obs()` does not refresh exteroception by\ndefault, and `env_info[\"visual_dict\"]` may therefore be empty or stale unless\nvisuals were explicitly updated at the current simulation time.\n\n## Backend and safety boundary\n\nBase MuJoCo model loading, stepping, state inspection, and the `MJRenderer`\noffscreen path are CPU\u002Fbase-package capabilities, subject to a usable graphics\nbackend for pixel rendering. Onscreen viewing additionally requires a display\nand native viewer support. MJX\u002FJAX\u002FCUDA is optional and is not established by\nan offscreen CPU check; route it to the MJX sub-skill and report it separately.\n\nBefore reporting success, check the actual artifact: model load\u002Fstep succeeded,\nrendered arrays have the requested dimensions, and every requested output file\nexists and is non-empty. For failure symptoms and recovery branches, use\n[troubleshooting.md](references\u002Ftroubleshooting.md).",{"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\u002Fmyosuite\u002Fsub-skills\u002Fsimulation-rendering",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 simulation-rendering",{"tool":40,"command":41},"bunx","bunx skills add VectorSpaceLab\u002FAREX-Skill --skill simulation-rendering",{"tool":43,"command":44},"pnpm","pnpm dlx skills add VectorSpaceLab\u002FAREX-Skill --skill simulation-rendering","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fmyosuite\u002Fsub-skills\u002Fsimulation-rendering\u002FSKILL.md",[47,53,57],{"path":48,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":51,"sourceHtmlUrl":52},"references\u002Frendering-api.md","REFERENCE","参考文档",8706,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fmyosuite\u002Fsub-skills\u002Fsimulation-rendering\u002Freferences\u002Frendering-api.md",{"path":54,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":55,"sourceHtmlUrl":56},"references\u002Ftroubleshooting.md",8366,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fmyosuite\u002Fsub-skills\u002Fsimulation-rendering\u002Freferences\u002Ftroubleshooting.md",{"path":58,"resourceType":59,"resourceTypeDesc":60,"sizeBytes":61,"sourceHtmlUrl":62},"scripts\u002Fcheck_mujoco_xml.py","SCRIPT","脚本",5881,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fmyosuite\u002Fsub-skills\u002Fsimulation-rendering\u002Fscripts\u002Fcheck_mujoco_xml.py",[64,71,78,85,92,99],{"skillNo":65,"name":66,"nameZh":6,"description":67,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":68,"categories":69,"keywords":70,"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":72,"name":73,"nameZh":6,"description":74,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":75,"categories":76,"keywords":77,"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":79,"name":80,"nameZh":6,"description":81,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":82,"categories":83,"keywords":84,"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":86,"name":87,"nameZh":6,"description":88,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":89,"categories":90,"keywords":91,"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":93,"name":94,"nameZh":6,"description":95,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":96,"categories":97,"keywords":98,"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":100,"name":101,"nameZh":6,"description":102,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":103,"categories":104,"keywords":105,"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",[],[]]