[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M791NF31G5CBXDSN4YAWQS4M":3,"skill-same:01M791NF31G5CBXDSN4YAWQS4M":71},{"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},"01M791NF31G5CBXDSN4YAWQS4M","advanced-graph-control",null,"Advanced graph-control workflows for ONNX Simplifier: custom rewriters, FunctionProto rules, metrics, graph diffs, metadata, and profiling.","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\u002Fonnxsim\u002Fsub-skills\u002Fadvanced-graph-control\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: advanced-graph-control\ndescription: \"Advanced graph-control workflows for ONNX Simplifier: custom\n  rewriters, FunctionProto rules, metrics, graph diffs, metadata, and\n  profiling.\"\ndisable-model-invocation: true\nmetadata:\n  disco-role: operating\nlicense: Apache 2.0\n---\n\n# advanced-graph-control\n\nUse this sub-skill when the task is not just \"simplify this ONNX model\", but\nneeds control over the simplification fixed point, graph rewrite rules, schema\nregistration, graph inspection, model metrics, metadata annotations, or profiling\ntraces.\n\n## Route first\n\n- Basic CLI\u002FPython simplification, input shapes, correctness checks, providers,\n  external data, target opset, constant folding, and ordinary optimizer skipping\n  belong in sibling sub-skill path `..\u002Fpython-simplification\u002FSKILL.md`.\n- Python wheel\u002Fsource builds, C API, Rust API, WASM\u002FNode\u002Fnpm, and binding build\n  commands belong in sibling sub-skill path `..\u002Fbindings-and-packaging\u002FSKILL.md`.\n- This sub-skill owns the semantics of FunctionProto rewrite rules; C\u002FRust users\n  can use the same rule bytes, but binding-specific invocation belongs to the\n  bindings sub-skill.\n\n## Primary references\n\n- Rewriter, FunctionProto, and custom schema contracts:\n  [`references\u002Frewriters-and-rules.md`](references\u002Frewriters-and-rules.md)\n- ModelInfo, MACs\u002FFLOPs\u002Fmemory metrics, graph diff, metadata, and profiling:\n  [`references\u002Fmetrics-and-profiling.md`](references\u002Fmetrics-and-profiling.md)\n- Advanced graph-control failure modes:\n  [`references\u002Ftroubleshooting.md`](references\u002Ftroubleshooting.md)\n\n## Bundled smoke scripts\n\nRun these from any environment where `onnx` and `onnxsim` are importable; they do\nnot require the original repository checkout.\n\n```bash\npython scripts\u002Ffunction_rule_smoke.py --help\npython scripts\u002Ffunction_rule_smoke.py\npython scripts\u002Fmodel_metrics_smoke.py --help\npython scripts\u002Fmodel_metrics_smoke.py\n```\n\n- [`scripts\u002Ffunction_rule_smoke.py`](scripts\u002Ffunction_rule_smoke.py) builds a\n  tiny `MatMul + Add` model, applies a pure-data FunctionProto rule that rewrites\n  it to `Gemm`, skips the built-in Gemm fusion so the rule is provably\n  responsible, and asserts the rewritten graph contains `Gemm`.\n- [`scripts\u002Fmodel_metrics_smoke.py`](scripts\u002Fmodel_metrics_smoke.py) builds a\n  tiny model, reports `ModelInfo` metrics, annotates metrics into\n  `metadata_props`, and prints a compact graph-diff summary.\n\n## Workflow: custom Python rewriter\n\nUse `custom_rewriter=` when the rewrite is Python-only or needs a richer matcher\nthan onnxsim's built-in FunctionProto matcher.\n\n1. Build or load an `onnx.ModelProto`.\n2. Define a callable that accepts `onnx.ModelProto` and either:\n   - returns a rewritten `onnx.ModelProto`,\n   - mutates the input model in place and returns `None`, or\n   - returns `False` when this fixed-point round changed nothing.\n3. Call `onnxsim.simplify(model, custom_rewriter=callable, check_n=...)`.\n4. Validate both the ONNX checker and, when safe, `check_n > 0`.\n5. If a rewriter introduces custom-domain or newer-opset ops, register schemas or\n   convert the model to the expected opset before simplification.\n\nPrefer returning `False` on no-op rounds; it is the no-change sentinel that\nprevents unnecessary serialize\u002Fparse copies and helps the fixed point converge.\n\n## Workflow: binding-portable FunctionProto rule\n\nUse `function_rewrite_rules=` when a rule can be expressed as pure ONNX graph\ndata and should also be usable from C or Rust.\n\n1. Express a `(pattern, replacement)` pair as `onnx.FunctionProto` objects,\n   usually with `onnx.parser.parse_function`.\n2. Treat pattern inputs as wildcards; pattern outputs are rewired to replacement\n   outputs.\n3. Use `@name` \u002F `ref_attr_name` for attribute wildcards that bind from the\n   matched node and substitute into the replacement.\n4. Pass `function_rewrite_rules=[(pattern, replacement)]` to `onnxsim.simplify`.\n5. If a built-in optimizer would perform the same rewrite, pass\n   `skipped_optimizers=[\"that_optimizer\"]` while validating the rule itself.\n\n`custom_rewriter` and `function_rewrite_rules` are mutually exclusive in one\n`simplify()` call.\n\n## Workflow: custom schemas\n\nIf the model contains custom ops that should remain schema-valid:\n\n```python\nimport onnx\nimport onnxsim\n\nonnx.defs.register_schema(my_op_schema)\n# simplify() imports registered schemas automatically by default.\nmodel_simp, ok = onnxsim.simplify(model)\n```\n\nCall `onnxsim.import_onnx_schemas()` explicitly when you need an import count or\nwant to prime the onnxsim registry. Disable the automatic import only when you\nknow the model does not require Python-registered schemas:\n`onnxsim.simplify(model, import_custom_schemas=False)` or CLI\n`--skip-schema-import`.\n\n## Workflow: metrics, metadata, graph diff\n\nUse `onnxsim.model_info` for static inspection without running inference:\n\n```python\nfrom onnxsim.model_info import ModelInfo, annotate_metadata, diff_graphs\n\ninfo = ModelInfo(model)\nannotated = annotate_metadata(model)\ndiff = diff_graphs(original_model, simplified_model)\n```\n\nMetrics include operator counts, serialized model size, MACs, FLOPs, memory\naccess, peak memory footprint, and compute density. Some values are symbolic\nwhen dynamic dimensions are named and `sympy` is installed; otherwise unresolved\ndimensions are best-effort lower bounds.\n\n## Workflow: profile simplification\n\nUse profiling when a simplification is unexpectedly slow, memory-heavy, or when\nconstant folding\u002Foptimizer cost attribution matters.\n\n```python\nmodel_simp, ok = onnxsim.simplify(model, profile=\"profile.json\")\nmodel_simp, ok = onnxsim.simplify(model, ort_profile=\"ort_profile\")\nmodel_simp, ok = onnxsim.simplify(\n    model, profile=\"profile.json\", merge_ort_profile=True\n)\n```\n\nCLI equivalents are `--profile`, `--ort-profile`, and `--merge-ort-profile`.\nOpen Chrome trace JSON files in `chrome:\u002F\u002Ftracing` or Perfetto. Use a scratch or\nuser-selected output directory for traces; do not leave large trace files in a\nruntime skill directory.\n\n## Validation checklist\n\n- Run the bundled `--help` checks before relying on a script in a new environment.\n- For rewrite rules, assert the intended op counts and run `onnx.checker`.\n- For semantic rewrites, use `check_n` or explicit `input_data`; use `check_n=0`\n  only for structural or intentionally semantics-changing transformations.\n- For metrics, document whether dynamic shapes made values symbolic or lower\n  bounds.\n- For profiling, confirm the expected trace files exist and that environment\n  variables used for profiling are not unintentionally left set after the call.\n","# advanced-graph-control\n\nUse this sub-skill when the task is not just \"simplify this ONNX model\", but\nneeds control over the simplification fixed point, graph rewrite rules, schema\nregistration, graph inspection, model metrics, metadata annotations, or profiling\ntraces.\n\n## Route first\n\n- Basic CLI\u002FPython simplification, input shapes, correctness checks, providers,\n  external data, target opset, constant folding, and ordinary optimizer skipping\n  belong in sibling sub-skill path `..\u002Fpython-simplification\u002FSKILL.md`.\n- Python wheel\u002Fsource builds, C API, Rust API, WASM\u002FNode\u002Fnpm, and binding build\n  commands belong in sibling sub-skill path `..\u002Fbindings-and-packaging\u002FSKILL.md`.\n- This sub-skill owns the semantics of FunctionProto rewrite rules; C\u002FRust users\n  can use the same rule bytes, but binding-specific invocation belongs to the\n  bindings sub-skill.\n\n## Primary references\n\n- Rewriter, FunctionProto, and custom schema contracts:\n  [`references\u002Frewriters-and-rules.md`](references\u002Frewriters-and-rules.md)\n- ModelInfo, MACs\u002FFLOPs\u002Fmemory metrics, graph diff, metadata, and profiling:\n  [`references\u002Fmetrics-and-profiling.md`](references\u002Fmetrics-and-profiling.md)\n- Advanced graph-control failure modes:\n  [`references\u002Ftroubleshooting.md`](references\u002Ftroubleshooting.md)\n\n## Bundled smoke scripts\n\nRun these from any environment where `onnx` and `onnxsim` are importable; they do\nnot require the original repository checkout.\n\n```bash\npython scripts\u002Ffunction_rule_smoke.py --help\npython scripts\u002Ffunction_rule_smoke.py\npython scripts\u002Fmodel_metrics_smoke.py --help\npython scripts\u002Fmodel_metrics_smoke.py\n```\n\n- [`scripts\u002Ffunction_rule_smoke.py`](scripts\u002Ffunction_rule_smoke.py) builds a\n  tiny `MatMul + Add` model, applies a pure-data FunctionProto rule that rewrites\n  it to `Gemm`, skips the built-in Gemm fusion so the rule is provably\n  responsible, and asserts the rewritten graph contains `Gemm`.\n- [`scripts\u002Fmodel_metrics_smoke.py`](scripts\u002Fmodel_metrics_smoke.py) builds a\n  tiny model, reports `ModelInfo` metrics, annotates metrics into\n  `metadata_props`, and prints a compact graph-diff summary.\n\n## Workflow: custom Python rewriter\n\nUse `custom_rewriter=` when the rewrite is Python-only or needs a richer matcher\nthan onnxsim's built-in FunctionProto matcher.\n\n1. Build or load an `onnx.ModelProto`.\n2. Define a callable that accepts `onnx.ModelProto` and either:\n   - returns a rewritten `onnx.ModelProto`,\n   - mutates the input model in place and returns `None`, or\n   - returns `False` when this fixed-point round changed nothing.\n3. Call `onnxsim.simplify(model, custom_rewriter=callable, check_n=...)`.\n4. Validate both the ONNX checker and, when safe, `check_n > 0`.\n5. If a rewriter introduces custom-domain or newer-opset ops, register schemas or\n   convert the model to the expected opset before simplification.\n\nPrefer returning `False` on no-op rounds; it is the no-change sentinel that\nprevents unnecessary serialize\u002Fparse copies and helps the fixed point converge.\n\n## Workflow: binding-portable FunctionProto rule\n\nUse `function_rewrite_rules=` when a rule can be expressed as pure ONNX graph\ndata and should also be usable from C or Rust.\n\n1. Express a `(pattern, replacement)` pair as `onnx.FunctionProto` objects,\n   usually with `onnx.parser.parse_function`.\n2. Treat pattern inputs as wildcards; pattern outputs are rewired to replacement\n   outputs.\n3. Use `@name` \u002F `ref_attr_name` for attribute wildcards that bind from the\n   matched node and substitute into the replacement.\n4. Pass `function_rewrite_rules=[(pattern, replacement)]` to `onnxsim.simplify`.\n5. If a built-in optimizer would perform the same rewrite, pass\n   `skipped_optimizers=[\"that_optimizer\"]` while validating the rule itself.\n\n`custom_rewriter` and `function_rewrite_rules` are mutually exclusive in one\n`simplify()` call.\n\n## Workflow: custom schemas\n\nIf the model contains custom ops that should remain schema-valid:\n\n```python\nimport onnx\nimport onnxsim\n\nonnx.defs.register_schema(my_op_schema)\n# simplify() imports registered schemas automatically by default.\nmodel_simp, ok = onnxsim.simplify(model)\n```\n\nCall `onnxsim.import_onnx_schemas()` explicitly when you need an import count or\nwant to prime the onnxsim registry. Disable the automatic import only when you\nknow the model does not require Python-registered schemas:\n`onnxsim.simplify(model, import_custom_schemas=False)` or CLI\n`--skip-schema-import`.\n\n## Workflow: metrics, metadata, graph diff\n\nUse `onnxsim.model_info` for static inspection without running inference:\n\n```python\nfrom onnxsim.model_info import ModelInfo, annotate_metadata, diff_graphs\n\ninfo = ModelInfo(model)\nannotated = annotate_metadata(model)\ndiff = diff_graphs(original_model, simplified_model)\n```\n\nMetrics include operator counts, serialized model size, MACs, FLOPs, memory\naccess, peak memory footprint, and compute density. Some values are symbolic\nwhen dynamic dimensions are named and `sympy` is installed; otherwise unresolved\ndimensions are best-effort lower bounds.\n\n## Workflow: profile simplification\n\nUse profiling when a simplification is unexpectedly slow, memory-heavy, or when\nconstant folding\u002Foptimizer cost attribution matters.\n\n```python\nmodel_simp, ok = onnxsim.simplify(model, profile=\"profile.json\")\nmodel_simp, ok = onnxsim.simplify(model, ort_profile=\"ort_profile\")\nmodel_simp, ok = onnxsim.simplify(\n    model, profile=\"profile.json\", merge_ort_profile=True\n)\n```\n\nCLI equivalents are `--profile`, `--ort-profile`, and `--merge-ort-profile`.\nOpen Chrome trace JSON files in `chrome:\u002F\u002Ftracing` or Perfetto. Use a scratch or\nuser-selected output directory for traces; do not leave large trace files in a\nruntime skill directory.\n\n## Validation checklist\n\n- Run the bundled `--help` checks before relying on a script in a new environment.\n- For rewrite rules, assert the intended op counts and run `onnx.checker`.\n- For semantic rewrites, use `check_n` or explicit `input_data`; use `check_n=0`\n  only for structural or intentionally semantics-changing transformations.\n- For metrics, document whether dynamic shapes made values symbolic or lower\n  bounds.\n- For profiling, confirm the expected trace files exist and that environment\n  variables used for profiling are not unintentionally left set after the call.\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\u002Fonnxsim\u002Fsub-skills\u002Fadvanced-graph-control",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 advanced-graph-control",{"tool":40,"command":41},"bunx","bunx skills add VectorSpaceLab\u002FAREX-Skill --skill advanced-graph-control",{"tool":43,"command":44},"pnpm","pnpm dlx skills add VectorSpaceLab\u002FAREX-Skill --skill advanced-graph-control","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fonnxsim\u002Fsub-skills\u002Fadvanced-graph-control\u002FSKILL.md",[47,53,57,61,67],{"path":48,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":51,"sourceHtmlUrl":52},"references\u002Fmetrics-and-profiling.md","REFERENCE","参考文档",10550,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fonnxsim\u002Fsub-skills\u002Fadvanced-graph-control\u002Freferences\u002Fmetrics-and-profiling.md",{"path":54,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":55,"sourceHtmlUrl":56},"references\u002Frewriters-and-rules.md",9794,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fonnxsim\u002Fsub-skills\u002Fadvanced-graph-control\u002Freferences\u002Frewriters-and-rules.md",{"path":58,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":59,"sourceHtmlUrl":60},"references\u002Ftroubleshooting.md",11291,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fonnxsim\u002Fsub-skills\u002Fadvanced-graph-control\u002Freferences\u002Ftroubleshooting.md",{"path":62,"resourceType":63,"resourceTypeDesc":64,"sizeBytes":65,"sourceHtmlUrl":66},"scripts\u002Ffunction_rule_smoke.py","SCRIPT","脚本",5790,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fonnxsim\u002Fsub-skills\u002Fadvanced-graph-control\u002Fscripts\u002Ffunction_rule_smoke.py",{"path":68,"resourceType":63,"resourceTypeDesc":64,"sizeBytes":69,"sourceHtmlUrl":70},"scripts\u002Fmodel_metrics_smoke.py",7550,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fonnxsim\u002Fsub-skills\u002Fadvanced-graph-control\u002Fscripts\u002Fmodel_metrics_smoke.py",[72,79,86,93,100,107],{"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},"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":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},"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":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},"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":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},"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":101,"name":102,"nameZh":6,"description":103,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":104,"categories":105,"keywords":106,"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":108,"name":109,"nameZh":6,"description":110,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":111,"categories":112,"keywords":113,"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",[],[]]