[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M792SHGZVZE8CFENRARMVBSF":3,"skill-same:01M792SHGZVZE8CFENRARMVBSF":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},"01M792SHGZVZE8CFENRARMVBSF","evaluation-reproduction",null,"Use PixelRAG evaluation scripts and paper reproduction harnesses safely, including reader\u002Fsearch serve preflights and graders.","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\u002Fpixelrag\u002Fsub-skills\u002Fevaluation-reproduction\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: evaluation-reproduction\ndescription: \"Use PixelRAG evaluation scripts and paper reproduction harnesses\n  safely, including reader\u002Fsearch serve preflights and graders.\"\ndisable-model-invocation: true\nmetadata:\n  disco-role: operating\nlicense: Apache 2.0\n---\n\n# PixelRAG Evaluation and Reproduction\n\nUse this sub-skill when the task asks to reproduce PixelRAG paper results, run `eval\u002F` benchmark cells, debug `reproduce.sh`, configure reader\u002Fsearch serves, compare retrieval modes, or understand grader\u002Foutput metadata.\n\n## Start Here\n\n1. Decide whether the user wants a **quick pipeline smoke** or a **paper-matching run**.\n   - Quick smoke can use a public search API or a tiny self-hosted index.\n   - Paper Table 1 reproduction requires specific search indexes, reader model, large tile\u002Findex data, and grader credentials.\n2. Validate services before running a benchmark. Use [pixelrag_eval_preflight.py](scripts\u002Fpixelrag_eval_preflight.py) or reproduce the shell preflight manually.\n3. Confirm credentials:\n   - OpenAI key\u002Fbase URL for LLM judge on NQ\u002FNQ-Tables and some VQA grading paths.\n   - Optional MiniMax key for MiniMax reader experiments.\n4. Record the run metadata stamped by the harness: dataset\u002Fsplit\u002Fn, reader, retrieval URL\u002Fstatus, top-k, instruction, and grader.\n\n## Read or Run\n\n- Read [paper-reproduction.md](references\u002Fpaper-reproduction.md) for `reproduce.sh`, ports, roles, expected resources, and public-API caveats.\n- Read [eval-api-reference.md](references\u002Feval-api-reference.md) for `run_bench.py`, retrieval flags, model config, output files, and graders.\n- Read [troubleshooting.md](references\u002Ftroubleshooting.md) for empty retrieval, wrong index, judge failures, query-image path issues, and slow on-demand rendering.\n- Run [pixelrag_eval_preflight.py](scripts\u002Fpixelrag_eval_preflight.py) before an expensive eval.\n\n## Common Routes\n\n| Request | Action |\n| --- | --- |\n| \"Reproduce Table 1 NQ base\" | Check reader `Qwen\u002FQwen3.5-4B`, base pixel serve vector count, OpenAI judge, then run the `nq\u002Fbase` cell. |\n| \"Try a public API smoke\" | Use direct `run_bench.py --local-api-url` mode; warn that public base endpoint may not match normed paper index. |\n| \"Why is score near zero?\" | Check preflight, retrieval timeout, grader key\u002Fbase URL, reader model, and whether closed-book fallback occurred. |\n| \"Run MiniMax reader\" | Use model config rules and set MiniMax endpoint\u002Fkey\u002Fcontext length explicitly. |\n| \"Need full training checkpoint eval\" | Route to `..\u002Ftraining-and-data\u002FSKILL.md` for training outputs, then return here for eval harness. |\n\n## Boundaries\n\n- Do not download hundreds of GB of indexes or tiles unless the user explicitly approves.\n- Do not launch vLLM, search serves, or long benchmark jobs silently.\n- Do not treat strict exact-match NQ\u002FNQ-Tables scores as paper numbers; the paper used an LLM judge for those cells.\n- Keep benchmark outputs and reports outside the runtime skill tree.\n","# PixelRAG Evaluation and Reproduction\n\nUse this sub-skill when the task asks to reproduce PixelRAG paper results, run `eval\u002F` benchmark cells, debug `reproduce.sh`, configure reader\u002Fsearch serves, compare retrieval modes, or understand grader\u002Foutput metadata.\n\n## Start Here\n\n1. Decide whether the user wants a **quick pipeline smoke** or a **paper-matching run**.\n   - Quick smoke can use a public search API or a tiny self-hosted index.\n   - Paper Table 1 reproduction requires specific search indexes, reader model, large tile\u002Findex data, and grader credentials.\n2. Validate services before running a benchmark. Use [pixelrag_eval_preflight.py](scripts\u002Fpixelrag_eval_preflight.py) or reproduce the shell preflight manually.\n3. Confirm credentials:\n   - OpenAI key\u002Fbase URL for LLM judge on NQ\u002FNQ-Tables and some VQA grading paths.\n   - Optional MiniMax key for MiniMax reader experiments.\n4. Record the run metadata stamped by the harness: dataset\u002Fsplit\u002Fn, reader, retrieval URL\u002Fstatus, top-k, instruction, and grader.\n\n## Read or Run\n\n- Read [paper-reproduction.md](references\u002Fpaper-reproduction.md) for `reproduce.sh`, ports, roles, expected resources, and public-API caveats.\n- Read [eval-api-reference.md](references\u002Feval-api-reference.md) for `run_bench.py`, retrieval flags, model config, output files, and graders.\n- Read [troubleshooting.md](references\u002Ftroubleshooting.md) for empty retrieval, wrong index, judge failures, query-image path issues, and slow on-demand rendering.\n- Run [pixelrag_eval_preflight.py](scripts\u002Fpixelrag_eval_preflight.py) before an expensive eval.\n\n## Common Routes\n\n| Request | Action |\n| --- | --- |\n| \"Reproduce Table 1 NQ base\" | Check reader `Qwen\u002FQwen3.5-4B`, base pixel serve vector count, OpenAI judge, then run the `nq\u002Fbase` cell. |\n| \"Try a public API smoke\" | Use direct `run_bench.py --local-api-url` mode; warn that public base endpoint may not match normed paper index. |\n| \"Why is score near zero?\" | Check preflight, retrieval timeout, grader key\u002Fbase URL, reader model, and whether closed-book fallback occurred. |\n| \"Run MiniMax reader\" | Use model config rules and set MiniMax endpoint\u002Fkey\u002Fcontext length explicitly. |\n| \"Need full training checkpoint eval\" | Route to `..\u002Ftraining-and-data\u002FSKILL.md` for training outputs, then return here for eval harness. |\n\n## Boundaries\n\n- Do not download hundreds of GB of indexes or tiles unless the user explicitly approves.\n- Do not launch vLLM, search serves, or long benchmark jobs silently.\n- Do not treat strict exact-match NQ\u002FNQ-Tables scores as paper numbers; the paper used an LLM judge for those cells.\n- Keep benchmark outputs and reports outside the runtime skill tree.\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\u002Fpixelrag\u002Fsub-skills\u002Fevaluation-reproduction",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 evaluation-reproduction",{"tool":40,"command":41},"bunx","bunx skills add VectorSpaceLab\u002FAREX-Skill --skill evaluation-reproduction",{"tool":43,"command":44},"pnpm","pnpm dlx skills add VectorSpaceLab\u002FAREX-Skill --skill evaluation-reproduction","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fpixelrag\u002Fsub-skills\u002Fevaluation-reproduction\u002FSKILL.md",[47,53,57,61],{"path":48,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":51,"sourceHtmlUrl":52},"references\u002Feval-api-reference.md","REFERENCE","参考文档",2700,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fpixelrag\u002Fsub-skills\u002Fevaluation-reproduction\u002Freferences\u002Feval-api-reference.md",{"path":54,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":55,"sourceHtmlUrl":56},"references\u002Fpaper-reproduction.md",2768,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fpixelrag\u002Fsub-skills\u002Fevaluation-reproduction\u002Freferences\u002Fpaper-reproduction.md",{"path":58,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":59,"sourceHtmlUrl":60},"references\u002Ftroubleshooting.md",2171,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fpixelrag\u002Fsub-skills\u002Fevaluation-reproduction\u002Freferences\u002Ftroubleshooting.md",{"path":62,"resourceType":63,"resourceTypeDesc":64,"sizeBytes":65,"sourceHtmlUrl":66},"scripts\u002Fpixelrag_eval_preflight.py","SCRIPT","脚本",2506,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fpixelrag\u002Fsub-skills\u002Fevaluation-reproduction\u002Fscripts\u002Fpixelrag_eval_preflight.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",[],[]]