[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M78WQ2V6VRATDYRVGRDW1VZ1":3,"skill-same:01M78WQ2V6VRATDYRVGRDW1VZ1":57},{"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},"01M78WQ2V6VRATDYRVGRDW1VZ1","filtered-transformations",null,"Use Equinox filtered JAX transformations over mixed PyTree models.","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\u002Fequinox\u002Fsub-skills\u002Ffiltered-transformations\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: filtered-transformations\ndescription: \"Use Equinox filtered JAX transformations over mixed PyTree models.\"\nmetadata:\n  disco-role: operating\ndisable-model-invocation: true\nlicense: Apache 2.0\n---\n\n# Filtered Transformations\n\nUse this sub-skill when applying JAX transforms to Equinox modules or other\nmixed PyTrees whose leaves are not all arrays.\n\n## Use when\n\n- A raw `jax.jit`, `jax.grad`, `jax.vmap`, or `jax.pmap` rejects an Equinox\n  model or callable PyTree.\n- The task names `filter_jit`, `filter_grad`, `filter_value_and_grad`,\n  `filter_vmap`, `filter_pmap`, `filter_eval_shape`, `filter_make_jaxpr`,\n  `filter_shard`, `filter_checkpoint`, or `filter_pure_callback`.\n- The user needs custom filtered derivatives with `filter_custom_jvp` or\n  `filter_custom_vjp`.\n- A training step must combine filtered gradients, Optax updates, and JIT.\n- The task involves pmap\u002Fsharding behavior for PyTrees with static leaves.\n\n## Route elsewhere\n\n- Use `..\u002Fmodule-and-trees\u002F` when the main task is defining fields,\n  partitioning a model manually, or performing tree surgery.\n- Use `..\u002Fnn-and-state\u002F` when the main task is layer construction, `BatchNorm`,\n  `Sequential`, `State`, or inference-mode toggles.\n- Use `..\u002Fdiagnostics-and-serialization\u002F` for runtime error\u002Fdebug behavior,\n  serialization, or pretty printing.\n- Use `..\u002Finternal-advanced\u002F` for `equinox.internal.noinline`, internal loops,\n  primitive authoring helpers, or other semi-public utilities.\n\n## Read first\n\n- [`references\u002Ffiltered-transformations.md`](references\u002Ffiltered-transformations.md) for API groups, patterns, and\n  examples.\n- [`references\u002Ftroubleshooting.md`](references\u002Ftroubleshooting.md) for transform-specific failure modes.\n- [`..\u002F..\u002Freferences\u002Fapi-reference.md`](..\u002F..\u002Freferences\u002Fapi-reference.md) for signature summaries.\n\n## Core workflow\n\n1. Determine which leaves should be traced as arrays. Equinox defaults usually\n   treat JAX\u002FNumPy arrays as dynamic and everything else as static.\n2. Replace raw JAX transforms with the corresponding `eqx.filter_*` transform\n   when a whole mixed PyTree crosses the transform boundary.\n3. Use `eqx.partition`\u002F`eqx.combine` only when you need explicit control or a\n   raw JAX transform API.\n4. For training steps, put loss, gradient computation, optimizer update, and\n   `eqx.apply_updates` inside one compiled function when possible.\n5. For batched models, remember that most `equinox.nn` layers act on one example;\n   use `jax.vmap` or `eqx.filter_vmap` around the call.\n6. For `filter_pmap` or `filter_shard`, check the device\u002Fsharding assumptions\n   before declaring backend coverage.\n\n## Minimal validation\n\nFrom the generated skill root (set `EQUINOX_SKILL_ROOT` to its absolute path),\nrun the bundled smoke helper after installing Equinox:\n\n```bash\ncd \"$EQUINOX_SKILL_ROOT\" && python scripts\u002Fsmoke.py --mode transformations\n```\n\nFor CPU multi-device `filter_pmap` smoke coverage:\n\n```bash\ncd \"$EQUINOX_SKILL_ROOT\" && python scripts\u002Fsmoke.py --mode transformations --two-cpu-devices\n```\n\nThe helper configures two logical CPU devices before importing JAX when that\nflag is supplied.\n\n## Scope note\n\nThis sub-skill distills Equinox's filtered transform behavior into the bundled\nreferences and smoke helper. The source repository and its test\u002Fdocs checkout\nwere construction evidence only, not runtime dependencies.\n## Key cautions\n\n- A CPU `filter_pmap` smoke validates API structure, not accelerator semantics.\n- `filter_grad` differentiates floating-point\u002Fcomplex array leaves in the first\n  argument and returns a PyTree matching that input; nondifferentiable\u002Fstatic\n  leaves appear as `None` gradients.\n- Donation options are explicit: `filter_jit` supports `donate=\"all\"`,\n  `\"all-except-first\"`, warning variants, and `\"none\"`; `filter_pmap` supports\n  `\"all\"`, `\"warn\"`, and `\"none\"`.\n- If a problem is actually about layer state or batch normalization axes, route\n  to `nn-and-state` after confirming the transform boundary.\n","# Filtered Transformations\n\nUse this sub-skill when applying JAX transforms to Equinox modules or other\nmixed PyTrees whose leaves are not all arrays.\n\n## Use when\n\n- A raw `jax.jit`, `jax.grad`, `jax.vmap`, or `jax.pmap` rejects an Equinox\n  model or callable PyTree.\n- The task names `filter_jit`, `filter_grad`, `filter_value_and_grad`,\n  `filter_vmap`, `filter_pmap`, `filter_eval_shape`, `filter_make_jaxpr`,\n  `filter_shard`, `filter_checkpoint`, or `filter_pure_callback`.\n- The user needs custom filtered derivatives with `filter_custom_jvp` or\n  `filter_custom_vjp`.\n- A training step must combine filtered gradients, Optax updates, and JIT.\n- The task involves pmap\u002Fsharding behavior for PyTrees with static leaves.\n\n## Route elsewhere\n\n- Use `..\u002Fmodule-and-trees\u002F` when the main task is defining fields,\n  partitioning a model manually, or performing tree surgery.\n- Use `..\u002Fnn-and-state\u002F` when the main task is layer construction, `BatchNorm`,\n  `Sequential`, `State`, or inference-mode toggles.\n- Use `..\u002Fdiagnostics-and-serialization\u002F` for runtime error\u002Fdebug behavior,\n  serialization, or pretty printing.\n- Use `..\u002Finternal-advanced\u002F` for `equinox.internal.noinline`, internal loops,\n  primitive authoring helpers, or other semi-public utilities.\n\n## Read first\n\n- [`references\u002Ffiltered-transformations.md`](references\u002Ffiltered-transformations.md) for API groups, patterns, and\n  examples.\n- [`references\u002Ftroubleshooting.md`](references\u002Ftroubleshooting.md) for transform-specific failure modes.\n- [`..\u002F..\u002Freferences\u002Fapi-reference.md`](..\u002F..\u002Freferences\u002Fapi-reference.md) for signature summaries.\n\n## Core workflow\n\n1. Determine which leaves should be traced as arrays. Equinox defaults usually\n   treat JAX\u002FNumPy arrays as dynamic and everything else as static.\n2. Replace raw JAX transforms with the corresponding `eqx.filter_*` transform\n   when a whole mixed PyTree crosses the transform boundary.\n3. Use `eqx.partition`\u002F`eqx.combine` only when you need explicit control or a\n   raw JAX transform API.\n4. For training steps, put loss, gradient computation, optimizer update, and\n   `eqx.apply_updates` inside one compiled function when possible.\n5. For batched models, remember that most `equinox.nn` layers act on one example;\n   use `jax.vmap` or `eqx.filter_vmap` around the call.\n6. For `filter_pmap` or `filter_shard`, check the device\u002Fsharding assumptions\n   before declaring backend coverage.\n\n## Minimal validation\n\nFrom the generated skill root (set `EQUINOX_SKILL_ROOT` to its absolute path),\nrun the bundled smoke helper after installing Equinox:\n\n```bash\ncd \"$EQUINOX_SKILL_ROOT\" && python scripts\u002Fsmoke.py --mode transformations\n```\n\nFor CPU multi-device `filter_pmap` smoke coverage:\n\n```bash\ncd \"$EQUINOX_SKILL_ROOT\" && python scripts\u002Fsmoke.py --mode transformations --two-cpu-devices\n```\n\nThe helper configures two logical CPU devices before importing JAX when that\nflag is supplied.\n\n## Scope note\n\nThis sub-skill distills Equinox's filtered transform behavior into the bundled\nreferences and smoke helper. The source repository and its test\u002Fdocs checkout\nwere construction evidence only, not runtime dependencies.\n## Key cautions\n\n- A CPU `filter_pmap` smoke validates API structure, not accelerator semantics.\n- `filter_grad` differentiates floating-point\u002Fcomplex array leaves in the first\n  argument and returns a PyTree matching that input; nondifferentiable\u002Fstatic\n  leaves appear as `None` gradients.\n- Donation options are explicit: `filter_jit` supports `donate=\"all\"`,\n  `\"all-except-first\"`, warning variants, and `\"none\"`; `filter_pmap` supports\n  `\"all\"`, `\"warn\"`, and `\"none\"`.\n- If a problem is actually about layer state or batch normalization axes, route\n  to `nn-and-state` after confirming the transform boundary.\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\u002Fequinox\u002Fsub-skills\u002Ffiltered-transformations",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 filtered-transformations",{"tool":40,"command":41},"bunx","bunx skills add VectorSpaceLab\u002FAREX-Skill --skill filtered-transformations",{"tool":43,"command":44},"pnpm","pnpm dlx skills add VectorSpaceLab\u002FAREX-Skill --skill filtered-transformations","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fequinox\u002Fsub-skills\u002Ffiltered-transformations\u002FSKILL.md",[47,53],{"path":48,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":51,"sourceHtmlUrl":52},"references\u002Ffiltered-transformations.md","REFERENCE","参考文档",5603,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fequinox\u002Fsub-skills\u002Ffiltered-transformations\u002Freferences\u002Ffiltered-transformations.md",{"path":54,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":55,"sourceHtmlUrl":56},"references\u002Ftroubleshooting.md",4809,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fequinox\u002Fsub-skills\u002Ffiltered-transformations\u002Freferences\u002Ftroubleshooting.md",[58,65,72,79,86,93],{"skillNo":59,"name":60,"nameZh":6,"description":61,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":62,"categories":63,"keywords":64,"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":66,"name":67,"nameZh":6,"description":68,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":69,"categories":70,"keywords":71,"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":73,"name":74,"nameZh":6,"description":75,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":76,"categories":77,"keywords":78,"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":80,"name":81,"nameZh":6,"description":82,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":83,"categories":84,"keywords":85,"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":87,"name":88,"nameZh":6,"description":89,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":90,"categories":91,"keywords":92,"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":94,"name":95,"nameZh":6,"description":96,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":97,"categories":98,"keywords":99,"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",[],[]]