[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6CC73VC7DQV1V5BYGET6FEV":3,"skill-same:01M6CC73VC7DQV1V5BYGET6FEV":83},{"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":38,"linkOnly":39,"markdownContent":40,"markdownBody":41,"frontMatter":42,"refName":52,"commitSha":53,"skillDirectory":54,"forksCount":55,"repositoryHtmlUrl":56,"repositoryHomepageUrl":6,"ownerHtmlUrl":57,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":58,"install":59,"resources":72},"01M6CC73VC7DQV1V5BYGET6FEV","meta-learn",null,"This skill should be used when the user asks to \"improve evaluator\", \"learn from evaluations\", \"evolve evaluation prompts\", \"check prediction accuracy\", \"review overridden verdicts\", \"meta-learn\", or \"how accurate are plugin evaluations\". Reviews past evaluations vs actual outcomes and evolves the screening\u002Freview prompts.","ray-manaloto\u002Fmacos-development-environment","ray-manaloto","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F227192835?v=4","01M6CBR7J13ZA01X5YB1A1QRJP","01M6CBR7P0S3NS8ZV81JYNST14",1,"MIT","https:\u002F\u002Fgithub.com\u002Fray-manaloto\u002Fmacos-development-environment\u002Fblob\u002Fmain\u002Frsm-subagents\u002Fplugins\u002Fmarketplace-evaluator\u002Fskills\u002Fmeta-learn\u002FSKILL.md",[],[18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37],"meta learn","learn","Meta","check","prompts","only","outcomes","user","edit","write","read","evolve","review","grep","improve","Bash","how","Evaluation","report","plugin","2026-11-10 10:12:30",false,"---\nname: meta-learn\ndescription: >\n  This skill should be used when the user asks to \"improve evaluator\", \"learn from evaluations\",\n  \"evolve evaluation prompts\", \"check prediction accuracy\", \"review overridden verdicts\",\n  \"meta-learn\", or \"how accurate are plugin evaluations\".\n  Reviews past evaluations vs actual outcomes and evolves the screening\u002Freview prompts.\nargument-hint: \"[--report-only] [--evolve-prompts]\"\nallowed-tools: [Read, Write, Edit, Glob, Grep, Bash, Agent]\n---\n\n# Meta-Learning for Marketplace Evaluator\n\nAnalyze past evaluation accuracy and evolve the evaluator's own prompts and criteria.\n\n## Data Sources\n\n1. **Evaluations JSON** — All past verdicts with confidence scores\n2. **Settings JSON** — Which plugins are actually enabled (ground truth)\n3. **Meta-learning JSON** — History of prompt evolution and accuracy metrics\n4. **User overrides** — Verdicts manually changed by the user\n\n## Analysis Pipeline\n\n### Step 1: Collect Outcomes\n\n1. Read current `.claude\u002Fsettings.json` to find all enabled plugins\n2. Compare against evaluation verdicts:\n   - INSTALL verdicts that ARE enabled → True Positive\n   - INSTALL verdicts that are NOT enabled → False Positive (user rejected)\n   - REJECT verdicts that ARE enabled → False Negative (user overrode)\n   - REJECT verdicts that are NOT enabled → True Negative\n3. Compute accuracy, precision, recall, F1\n\n### Step 2: Analyze Patterns\n\n1. For False Positives: Why did the evaluator recommend plugins the user rejected?\n   - Policy violations missed?\n   - Overlap with existing tooling not detected?\n   - Context budget impact underestimated?\n2. For False Negatives: Why did the evaluator reject plugins the user installed?\n   - Policy too strict? (user accepted a policy violation)\n   - Value underestimated?\n   - Goal alignment missed?\n\n### Step 3: Evolve Prompts (if `--evolve-prompts`)\n\n1. Spawn **meta-learner** agent (model from config)\n2. Agent reviews the pattern analysis and current agent prompts\n3. Agent proposes specific prompt modifications:\n   - Screening criteria adjustments\n   - Policy strictness calibration\n   - Goal keyword additions\n   - Confidence scoring weight changes\n4. Write proposals to meta-learning JSON for human review\n5. Do NOT auto-apply prompt changes — present for approval\n\n### Step 4: Report\n\nOutput a meta-learning report:\n\n```\nMeta-Learning Report\n====================\nEvaluations analyzed: 51\nAccuracy: 82% | Precision: 75% | Recall: 90%\n\nFalse Positives (recommended but not installed): 5\n  - plugin-x: User rejected because [pattern]\n  - plugin-y: Overlap with existing tooling\n\nFalse Negatives (rejected but user installed): 2\n  - plugin-z: Policy override accepted by user\n\nPrompt Evolution Proposals:\n  1. Relax bash-hooks policy for guard utilities (3 FN from this)\n  2. Add \"worktree\" keyword to agent-improvement goal\n  3. Increase overlap detection weight from 0.2 to 0.3\n```\n\n## Additional Resources\n\n- **`references\u002Fmeta-learning-patterns.md`** — Common patterns in evaluation drift and correction strategies\n","# Meta-Learning for Marketplace Evaluator\n\nAnalyze past evaluation accuracy and evolve the evaluator's own prompts and criteria.\n\n## Data Sources\n\n1. **Evaluations JSON** — All past verdicts with confidence scores\n2. **Settings JSON** — Which plugins are actually enabled (ground truth)\n3. **Meta-learning JSON** — History of prompt evolution and accuracy metrics\n4. **User overrides** — Verdicts manually changed by the user\n\n## Analysis Pipeline\n\n### Step 1: Collect Outcomes\n\n1. Read current `.claude\u002Fsettings.json` to find all enabled plugins\n2. Compare against evaluation verdicts:\n   - INSTALL verdicts that ARE enabled → True Positive\n   - INSTALL verdicts that are NOT enabled → False Positive (user rejected)\n   - REJECT verdicts that ARE enabled → False Negative (user overrode)\n   - REJECT verdicts that are NOT enabled → True Negative\n3. Compute accuracy, precision, recall, F1\n\n### Step 2: Analyze Patterns\n\n1. For False Positives: Why did the evaluator recommend plugins the user rejected?\n   - Policy violations missed?\n   - Overlap with existing tooling not detected?\n   - Context budget impact underestimated?\n2. For False Negatives: Why did the evaluator reject plugins the user installed?\n   - Policy too strict? (user accepted a policy violation)\n   - Value underestimated?\n   - Goal alignment missed?\n\n### Step 3: Evolve Prompts (if `--evolve-prompts`)\n\n1. Spawn **meta-learner** agent (model from config)\n2. Agent reviews the pattern analysis and current agent prompts\n3. Agent proposes specific prompt modifications:\n   - Screening criteria adjustments\n   - Policy strictness calibration\n   - Goal keyword additions\n   - Confidence scoring weight changes\n4. Write proposals to meta-learning JSON for human review\n5. Do NOT auto-apply prompt changes — present for approval\n\n### Step 4: Report\n\nOutput a meta-learning report:\n\n```\nMeta-Learning Report\n====================\nEvaluations analyzed: 51\nAccuracy: 82% | Precision: 75% | Recall: 90%\n\nFalse Positives (recommended but not installed): 5\n  - plugin-x: User rejected because [pattern]\n  - plugin-y: Overlap with existing tooling\n\nFalse Negatives (rejected but user installed): 2\n  - plugin-z: Policy override accepted by user\n\nPrompt Evolution Proposals:\n  1. Relax bash-hooks policy for guard utilities (3 FN from this)\n  2. Add \"worktree\" keyword to agent-improvement goal\n  3. Increase overlap detection weight from 0.2 to 0.3\n```\n\n## Additional Resources\n\n- **`references\u002Fmeta-learning-patterns.md`** — Common patterns in evaluation drift and correction strategies\n",{"name":5,"description":43,"allowed-tools":44,"argument-hint":51},"This skill should be used when the user asks to \"improve evaluator\", \"learn from evaluations\", \"evolve evaluation prompts\", \"check prediction accuracy\", \"review overridden verdicts\", \"meta-learn\", or \"how accurate are plugin evaluations\". Reviews past evaluations vs actual outcomes and evolves the screening\u002Freview prompts.\n",[45,46,47,48,49,33,50],"Read","Write","Edit","Glob","Grep","Agent","[--report-only] [--evolve-prompts]","main","716b17df3ef6719d88aeac180015008b7ddfff6f","rsm-subagents\u002Fplugins\u002Fmarketplace-evaluator\u002Fskills\u002Fmeta-learn",0,"https:\u002F\u002Fgithub.com\u002Fray-manaloto\u002Fmacos-development-environment","https:\u002F\u002Fgithub.com\u002Fray-manaloto","2026-08-03 16:55:50",{"source":8,"skillName":5,"commands":60,"prompt":70,"zipAvailable":71},[61,64,67],{"tool":62,"command":63},"npx","npx skills add ray-manaloto\u002Fmacos-development-environment --skill meta-learn",{"tool":65,"command":66},"bunx","bunx skills add ray-manaloto\u002Fmacos-development-environment --skill meta-learn",{"tool":68,"command":69},"pnpm","pnpm dlx skills add ray-manaloto\u002Fmacos-development-environment --skill meta-learn","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fray-manaloto\u002Fmacos-development-environment\u002Fblob\u002Fmain\u002Frsm-subagents\u002Fplugins\u002Fmarketplace-evaluator\u002Fskills\u002Fmeta-learn\u002FSKILL.md",true,[73,79],{"path":74,"resourceType":75,"resourceTypeDesc":76,"sizeBytes":77,"sourceHtmlUrl":78},"agents\u002Fopenai.yaml","REFERENCE","参考文档",116,"https:\u002F\u002Fgithub.com\u002Fmalinskibeniamin\u002Fskills\u002Fblob\u002Fmain\u002Fcodex-skills\u002Fproduct\u002Fagents\u002Fopenai.yaml",{"path":80,"resourceType":75,"resourceTypeDesc":76,"sizeBytes":81,"sourceHtmlUrl":82},"references\u002Fmeta-learning-patterns.md",2754,"https:\u002F\u002Fgithub.com\u002Fray-manaloto\u002Fmacos-development-environment\u002Fblob\u002Fmain\u002Frsm-subagents\u002Fplugins\u002Fmarketplace-evaluator\u002Fskills\u002Fmeta-learn\u002Freferences\u002Fmeta-learning-patterns.md",[84,106,123,141,161,175],{"skillNo":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":95,"publishedAt":38},"01M6CC8DXNJ39YWERYTWCGGCGG","next-workflow-step","Use this skill when the user wants to transition to the next workflow step, wrap up a session cleanly, prepare a handoff, or says things like \"next step\", \"continue the workflow\", \"wrap up\", \"hand off\", \"move to the next phase\", \"what's next in embrace\", or \"prepare for \u002Fclear\". Also triggers when the current octo:embrace phase appears complete and no explicit next action is set.","https:\u002F\u002Fgithub.com\u002Fray-manaloto\u002Fmacos-development-environment\u002Fblob\u002Fmain\u002Frsm-subagents\u002Fplugins\u002Fworkflow-toolkit\u002Fskills\u002Fnext-workflow-step\u002FSKILL.md",[90],{"categoryNo":91,"categoryName":92,"taxonomyType":93,"taxonomyTypeDesc":94},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[96,97,98,99,100,25,101,102,103,104,105],"next workflow step","workflow","up","use","session","wrap up","continue","handoff","prepare","SET",{"skillNo":107,"name":108,"nameZh":6,"description":109,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":110,"categories":111,"keywords":112,"publishedAt":38},"01M6CC8BRZBQHEBXGMKPDM422P","source-discovery","This skill should be used when the user asks to \"find sources\", \"discover URLs\", \"catalog a source\", \"classify a URL\", \"search for libraries\", \"check what exists\", or mentions source discovery, URL cataloging, source classification, or research source management.","https:\u002F\u002Fgithub.com\u002Fray-manaloto\u002Fmacos-development-environment\u002Fblob\u002Fmain\u002Frsm-subagents\u002Fplugins\u002Fresearch-review-toolkit\u002Fskills\u002Fsource-discovery\u002FSKILL.md",[],[113,114,115,116,117,118,119,21,120,25,121,122],"discovery","source discovery","research","libraries","search","classification","find","management","catalog","find sources",{"skillNo":124,"name":125,"nameZh":6,"description":126,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":127,"categories":128,"keywords":129,"publishedAt":38},"01M6CC89CGJP69A0PS22G2H1FV","review-synthesis","This skill should be used when the user asks to \"summarize review findings\", \"prioritize fixes\", \"rank review issues\", \"combine reviewer feedback\", \"create action items from review\", or mentions review synthesis, finding prioritization, severity ranking, or review result summarization.","https:\u002F\u002Fgithub.com\u002Fray-manaloto\u002Fmacos-development-environment\u002Fblob\u002Fmain\u002Frsm-subagents\u002Fplugins\u002Fresearch-review-toolkit\u002Fskills\u002Freview-synthesis\u002FSKILL.md",[],[30,130,131,132,133,134,135,25,136,137,138,139,140],"review synthesis","Synthesis","prioritize","create action","feedback","reviewer","summarize","prioritization","create","issues","severity",{"skillNo":142,"name":143,"nameZh":6,"description":144,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":145,"categories":146,"keywords":150,"publishedAt":38},"01M6CC872EZ9GRJY0HRCG0RT2M","research-scoring","This skill should be used when the user asks to \"score research\", \"check research quality\", \"evaluate coverage\", \"measure improvement\", \"assess research completeness\", or mentions research scoring, quality metrics, coverage gaps, improvement baseline, or research pipeline status.","https:\u002F\u002Fgithub.com\u002Fray-manaloto\u002Fmacos-development-environment\u002Fblob\u002Fmain\u002Frsm-subagents\u002Fplugins\u002Fresearch-review-toolkit\u002Fskills\u002Fresearch-scoring\u002FSKILL.md",[147],{"categoryNo":148,"categoryName":149,"taxonomyType":93,"taxonomyTypeDesc":94},"01JBSEEDCAT000000000000021","资料检索与研究",[151,115,152,153,154,155,156,157,21,25,158,159,160],"research scoring","status","coverage gaps","metrics","pipeline","baseline","quality","coverage","pipeline status","research quality",{"skillNo":162,"name":163,"nameZh":6,"description":164,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":165,"categories":166,"keywords":167,"publishedAt":38},"01M6CC852HSDS1FE4EJAT0RHGR","notebooklm-integration","This skill should be used when the user asks to \"ingest into NotebookLM\", \"query NotebookLM\", \"add sources to notebook\", \"cross-source synthesis\", \"check stale sources\", or mentions NotebookLM, notebook integration, source ingestion, or knowledge consolidation.","https:\u002F\u002Fgithub.com\u002Fray-manaloto\u002Fmacos-development-environment\u002Fblob\u002Fmain\u002Frsm-subagents\u002Fplugins\u002Fresearch-review-toolkit\u002Fskills\u002Fnotebooklm-integration\u002FSKILL.md",[],[168,169,170,171,21,25,172,173,174,131],"notebooklm integration","notebooklm","integration","knowledge","ingestion","ingest","stale",{"skillNo":176,"name":177,"nameZh":6,"description":178,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":179,"categories":180,"keywords":182,"publishedAt":38},"01M6CC82Z4V3205KCJWQZPS2VG","evidence-synthesis","This skill should be used when the user asks to \"synthesize findings\", \"combine research\", \"cross-reference sources\", \"summarize evidence\", \"write a deep review\", or mentions evidence synthesis, finding consolidation, cross-source analysis, or research summarization.","https:\u002F\u002Fgithub.com\u002Fray-manaloto\u002Fmacos-development-environment\u002Fblob\u002Fmain\u002Frsm-subagents\u002Fplugins\u002Fresearch-review-toolkit\u002Fskills\u002Fevidence-synthesis\u002FSKILL.md",[181],{"categoryNo":148,"categoryName":149,"taxonomyType":93,"taxonomyTypeDesc":94},[183,131,184,136,115,185,186,27,25,187,30],"evidence synthesis","evidence","deep review","analysis","reference"]