[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"extension-skill-yanhongxi-openclaw-proactive-self-improving-agent-es":3,"guides-for-yanhongxi-openclaw-proactive-self-improving-agent":219,"similar-k171wbhkefsa1s7m82mat9azn9867e51":220},{"_creationTime":4,"_id":5,"children":6,"community":7,"display":9,"evaluation":20,"identity":187,"isFallback":191,"parentExtension":192,"providers":193,"relations":198,"repo":200,"workflow":216},1778054617416.4724,"k171wbhkefsa1s7m82mat9azn9867e51",[],{"reviewCount":8},0,{"description":10,"installMethods":11,"name":12,"sourceUrl":13,"tags":14},"自动捕获经验并安全进化的技能。触发条件：(1)命令/操作失败时→记ERRORS.md (2)被用户纠正('不对'/'应该是')时→记LEARNINGS.md (3)用户需要不存在的能力时→记FEATURE_REQUESTS.md (4)外部API/工具出错时→记ERRORS.md (5)发现自己知识过时/错误时→记LEARNINGS.md (6)发现更好做法时→记LEARNINGS.md (7)每个任务完成时→回顾过程，有新经验则记LEARNINGS.md。去重原则：如果没有新经验或已有条目已覆盖则跳过不写。每次写入同时在.learnings/CHANGELOG.md追加JSONL日志。经验反复出现≥3次时晋升到AGENTS.md/TOOLS.md/SOUL.md。详见正文。",{},"Proactive Self-Improving Agent","https://github.com/yanhongxi-openclaw/proactive-self-improving-agent",[15,16,17,18,19],"agent","learning","self-improvement","logging","experience-capture",{"_creationTime":21,"_id":22,"extensionId":5,"locale":23,"result":24,"trustSignals":176,"workflow":185},1778054631760.5212,"kn7bz95chgbag4s2ma00z985ps867fem","en",{"checks":25,"evaluatedAt":166,"extensionSummary":167,"promptVersionExtension":168,"promptVersionScoring":169,"rationale":170,"score":171,"summary":172,"tags":173,"targetMarket":174,"tier":175},[26,31,34,37,41,44,49,53,56,59,63,67,70,74,77,80,83,86,89,92,96,100,104,108,112,115,118,121,125,128,131,134,137,140,144,147,150,153,156,159,163],{"category":27,"check":28,"severity":29,"summary":30},"Practical Utility","Problem relevance","pass","The description clearly states the problem of an agent needing to automatically capture experience and evolve safely, outlining specific triggers for learning and error logging.",{"category":27,"check":32,"severity":29,"summary":33},"Unique selling proposition","The extension offers a unique self-improving capability for agents by structuring experience capture and evolution, going beyond a simple wrapper by implementing a system for learning and refinement.",{"category":27,"check":35,"severity":29,"summary":36},"Production readiness","The skill provides a complete lifecycle for agent learning, from error capture and correction to experience promotion and skill extraction, with clear guidelines and logging.",{"category":38,"check":39,"severity":29,"summary":40},"Scope","Single responsibility principle","The extension's scope is focused on the self-improvement and experience logging of an agent, aligning with its name and description without branching into unrelated domains.",{"category":38,"check":42,"severity":29,"summary":43},"Description quality","The displayed description accurately reflects the skill's functionality, detailing its triggers, logging mechanisms, and evolutionary paths for agent self-improvement.",{"category":45,"check":46,"severity":47,"summary":48},"Invocation","Scoped tools","not_applicable","This extension does not expose specific tools; its functionality is integrated into the agent's core behavior through the SKILL.md instructions.",{"category":50,"check":51,"severity":47,"summary":52},"Documentation","Configuration & parameter reference","No configurable parameters or environment variables are used by this skill; its behavior is dictated by the instructions in SKILL.md and the structure of the .learnings directory.",{"category":38,"check":54,"severity":47,"summary":55},"Tool naming","This extension does not define explicit tools with user-facing names.",{"category":38,"check":57,"severity":47,"summary":58},"Minimal I/O surface","The extension does not expose specific tools with input/output schemas; its I/O is managed through file system interactions and agent internal state.",{"category":60,"check":61,"severity":29,"summary":62},"License","License usability","The extension is licensed under MIT, a permissive open-source license, clearly stated in the README.md.",{"category":64,"check":65,"severity":47,"summary":66},"Maintenance","Commit recency","The repository has no commit history available for evaluation.",{"category":64,"check":68,"severity":47,"summary":69},"Dependency Management","The extension has no third-party dependencies.",{"category":71,"check":72,"severity":47,"summary":73},"Security","Secret Management","This skill does not handle or manage secrets.",{"category":71,"check":75,"severity":29,"summary":76},"Injection","The skill's logic is contained within SKILL.md and does not involve loading external, untrusted data as executable code.",{"category":71,"check":78,"severity":29,"summary":79},"Transitive Supply-Chain Grenades","The extension does not fetch remote content at runtime or include external scripts, with all logic contained within the committed files.",{"category":71,"check":81,"severity":29,"summary":82},"Sandbox Isolation","The skill's operations are limited to interacting with the .learnings directory and the agent's internal state, without touching external files or paths.",{"category":71,"check":84,"severity":29,"summary":85},"Sandbox escape primitives","No detached process spawns or deny-retry loops are present in the skill's instructions.",{"category":71,"check":87,"severity":29,"summary":88},"Data Exfiltration","The skill does not make any outbound network calls or reference confidential data. All interactions are local to the agent's environment.",{"category":71,"check":90,"severity":29,"summary":91},"Hidden Text Tricks","Bundled files do not contain any hidden-steering tricks, invisible Unicode characters, or obfuscated content.",{"category":93,"check":94,"severity":29,"summary":95},"Hooks","Opaque code execution","The skill's logic is written in human-readable markdown (SKILL.md) and does not involve obfuscated code, base64 payloads, or runtime script fetching.",{"category":97,"check":98,"severity":29,"summary":99},"Portability","Structural Assumption","The skill relies on the presence of a `.learnings` directory, which is a documented precondition and explicitly created by the user, avoiding assumptions about other project structures.",{"category":101,"check":102,"severity":47,"summary":103},"Trust","Issues Attention","No GitHub issues data is available for evaluation.",{"category":105,"check":106,"severity":29,"summary":107},"Versioning","Release Management","The skill declares version 1.0.0 in its SKILL.md frontmatter.",{"category":109,"check":110,"severity":47,"summary":111},"Code Execution","Validation","The skill does not interact with external tools or accept structured input that requires validation via a schema library.",{"category":71,"check":113,"severity":29,"summary":114},"Unguarded Destructive Operations","The skill's operations are confined to file logging and state management within the .learnings directory, which are not considered destructive.",{"category":109,"check":116,"severity":29,"summary":117},"Error Handling","The SKILL.md provides clear instructions on how errors are logged and handled, indicating a fail-closed behavior for unexpected states and referencing error reporting mechanisms.",{"category":109,"check":119,"severity":29,"summary":120},"Logging","The skill explicitly logs actions and outcomes to .learnings/CHANGELOG.md in a structured JSONL format, serving as an audit trail.",{"category":122,"check":123,"severity":29,"summary":124},"Compliance","GDPR","The skill operates locally and does not process or transmit personal data.",{"category":122,"check":126,"severity":29,"summary":127},"Target market","The skill is designed for agent self-improvement and learning, with no specific regional or jurisdictional logic, making it globally applicable.",{"category":97,"check":129,"severity":29,"summary":130},"Runtime stability","The skill relies on standard file system operations and markdown instructions, ensuring cross-platform compatibility without assuming specific shells or runtimes.",{"category":45,"check":132,"severity":29,"summary":133},"Precise Purpose","The SKILL.md clearly defines the skill's purpose of agent self-improvement and experience capture, along with specific triggers and boundaries.",{"category":45,"check":135,"severity":29,"summary":136},"Concise Frontmatter","The frontmatter in SKILL.md is concise, effectively summarizing the core capability and providing trigger phrases within the character limit.",{"category":50,"check":138,"severity":29,"summary":139},"Concise Body","The SKILL.md is well-structured and under 500 lines, delegating detailed procedures to external files as needed.",{"category":141,"check":142,"severity":29,"summary":143},"Context","Progressive Disclosure","The SKILL.md outlines the overall flow and uses relative paths to link to detailed explanations or procedures, demonstrating progressive disclosure.",{"category":141,"check":145,"severity":47,"summary":146},"Forked exploration","The skill's function is not primarily deep exploration; it focuses on logging and learning based on agent actions rather than intensive code review or research.",{"category":27,"check":148,"severity":29,"summary":149},"Usage examples","The README provides clear installation and usage instructions, including the necessary command to create the .learnings directory, demonstrating a basic but functional example.",{"category":27,"check":151,"severity":29,"summary":152},"Edge cases","The SKILL.md details various failure modes and recovery paths, such as handling command failures, user corrections, and API errors, along with a distinct process for task completion reviews.",{"category":109,"check":154,"severity":47,"summary":155},"Tool Fallback","The skill does not rely on external tools like MCP servers; its functionality is self-contained within the agent's environment.",{"category":97,"check":157,"severity":29,"summary":158},"Stack assumptions","The skill assumes a standard agent runtime environment capable of executing markdown instructions and interacting with the file system, with no specific OS or shell dependencies.",{"category":160,"check":161,"severity":29,"summary":162},"Safety","Halt on unexpected state","The SKILL.md explicitly instructs the agent to halt and report on unexpected pre-state conditions, such as command failures or errors, ensuring a fail-closed behavior.",{"category":97,"check":164,"severity":29,"summary":165},"Cross-skill coupling","The skill is self-contained and does not implicitly rely on other skills, clearly defining its own scope for experience capture and learning.",1778054631479,"This skill enhances agent capabilities by automatically logging errors, user corrections, and task reviews into structured files (.learnings/LEARNINGS.md, ERRORS.md, FEATURE_REQUESTS.md). It includes mechanisms for experience evolution, such as promoting learnings to permanent rules or extracting them into new skills, all while maintaining safety and logging all actions in CHANGELOG.md.","2.0.0","3.4.0","The extension demonstrates excellent documentation quality, clear scope, and robust safety features. It successfully implements a sophisticated self-improvement system for agents, with clear instructions, logging, and versioning. No critical or warning findings were observed.",98,"A high-quality skill that enables agents to automatically capture and learn from their experiences, fostering safe self-improvement.",[15,16,17,18,19],"global","verified",{"codeQuality":177,"collectedAt":178,"documentation":179,"maintenance":181,"security":182,"testCoverage":184},{},1778054619509,{"descriptionLength":180,"readmeSize":8},342,{},{"hasNpmPackage":183,"smitheryVerified":183},false,{"hasCi":183,"hasTests":183},{"updatedAt":186},1778054631760,{"githubOwner":188,"githubRepo":189,"locale":23,"slug":189,"type":190},"yanhongxi-openclaw","proactive-self-improving-agent","skill",true,null,{"extract":194,"llm":197},{"commitSha":195,"license":196},"5af957709186392fb9892382d0ffb26e8095ad65","MIT",{"promptVersionExtension":168,"promptVersionScoring":169,"score":171,"targetMarket":174,"tier":175},{"repoId":199},"kd78yyb7ypqjg4av4r3p0feyfd864sdm",{"_creationTime":201,"_id":199,"identity":202,"providers":203,"workflow":213},1777995558409.8828,{"githubOwner":188,"githubRepo":189,"sourceUrl":13},{"discover":204,"github":207},{"sources":205},[206],"skills-sh",{"closedIssues90d":8,"forks":208,"openIssues90d":8,"pushedAt":209,"readmeSize":210,"stars":211,"topics":212},1,1772431289000,1171,10,[],{"discoverAt":214,"extractAt":215,"githubAt":215,"updatedAt":215},1777995558409,1778054618247,{"anyEnrichmentAt":217,"extractAt":218,"githubAt":217,"llmAt":186,"updatedAt":186},1778054617695,1778054617416,[],[221,249,278,307,336,365],{"_creationTime":222,"_id":223,"community":224,"display":225,"identity":235,"providers":238,"relations":243,"workflow":245},1778054243695.7876,"k179a82fbn3mx6xcy0qsx6hvr9866amn",{"reviewCount":8},{"description":226,"installMethods":227,"name":228,"sourceUrl":229,"tags":230},"Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Claude ('No, that's wrong...', 'Actually...'), (3) User requests a capability that doesn't exist, (4) An external API or tool fails, (5) Claude realizes its knowledge is outdated or incorrect, (6) A better approach is discovered for a recurring task. Also review learnings before major tasks.",{},"Self-Improvement","https://github.com/pskoett/self-improving-agent",[16,18,231,232,233,234],"error-handling","continuous-improvement","agent-productivity","hooks",{"githubOwner":236,"githubRepo":237,"locale":23,"slug":17,"type":190},"pskoett","self-improving-agent",{"extract":239,"llm":241},{"commitSha":240,"license":47},"6add06bdb330f4b967d7e2d9d7939d4aceea86ad",{"promptVersionExtension":168,"promptVersionScoring":169,"score":242,"targetMarket":174,"tier":175},92,{"repoId":244},"kd7485b926qh05ywgj7h6r4sdd865anj",{"anyEnrichmentAt":246,"extractAt":247,"githubAt":246,"llmAt":248,"updatedAt":248},1778054243980,1778054243695,1778054255258,{"_creationTime":250,"_id":251,"community":252,"display":253,"identity":264,"providers":268,"relations":272,"workflow":274},1778054432207.5422,"k178t7t7ek2kr53kqyz2tn8m4x867gmh",{"reviewCount":8},{"description":254,"installMethods":255,"name":256,"sourceUrl":257,"tags":258},"Manage OpenClaw bot configuration - channels, agents, security, and autopilot settings",{},"OpenClaw Config","https://github.com/adisinghstudent/easyclaw/tree/HEAD/skills/openclaw-config",[259,15,260,261,262,263],"openclaw","configuration","troubleshooting","typescript","devops",{"githubOwner":265,"githubRepo":266,"locale":23,"slug":267,"type":190},"adisinghstudent","easyclaw","openclaw-config",{"extract":269,"llm":271},{"commitSha":270,"license":196},"2a938af22323bcc822625828f6303cb75690e7c7",{"promptVersionExtension":168,"promptVersionScoring":169,"score":171,"targetMarket":174,"tier":175},{"repoId":273},"kd7fy171w716bjdays4yawdj91865292",{"anyEnrichmentAt":275,"extractAt":276,"githubAt":275,"llmAt":277,"updatedAt":277},1778054432549,1778054432207,1778054457102,{"_creationTime":279,"_id":280,"community":281,"display":282,"identity":292,"providers":295,"relations":300,"workflow":302},1777995614337.9856,"k177fpj38ebtn9tkr8x4kx39xd864dvz",{"reviewCount":8},{"description":283,"name":284,"sourceUrl":285,"tags":286},"TypeScript client for the Tandem autonomous agent engine HTTP + SSE API","Tandem TypeScript Client","https://github.com/frumu-ai/tandem/tree/HEAD/packages/tandem-client-ts",[262,287,288,289,15,290,291],"client","api","tandem","sdk","node-js",{"githubOwner":293,"githubRepo":289,"locale":23,"slug":294,"type":190},"frumu-ai","frumu-tandem-client",{"extract":296,"llm":298,"smithery":299},{"commitSha":297,"license":196},"25c8ae7c2f8b20d576a2b8d44946ac2e0ff9408d",{"promptVersionExtension":168,"promptVersionScoring":169,"score":171,"targetMarket":174,"tier":175},{"qualityScore":8,"totalActivations":8,"uniqueUsers":8,"useCount":8,"verified":183},{"repoId":301},"kd72pd305er4m7vecv9p6mnd4s865kea",{"anyEnrichmentAt":303,"extractAt":304,"githubAt":305,"llmAt":306,"smitheryAt":303,"updatedAt":306},1777995723550,1777995614338,1777995614791,1777995897177,{"_creationTime":308,"_id":309,"community":310,"display":311,"identity":323,"providers":326,"relations":330,"workflow":332},1778054812528.7205,"k179w7m3n2zkvrbhmmwn1stces86799q",{"reviewCount":8},{"description":312,"installMethods":313,"name":314,"sourceUrl":315,"tags":316},"3D web graphics with Three.js (WebGL/WebGPU). Capabilities: scenes, cameras, geometries, materials, lights, animations, model loading (GLTF/FBX), PBR materials, shadows, post-processing (bloom, SSAO, SSR), custom shaders, instancing, LOD, physics, VR/XR. Actions: create, build, animate, render 3D scenes/models. Keywords: Three.js, WebGL, WebGPU, 3D graphics, scene, camera, geometry, material, light, animation, GLTF, FBX, OrbitControls, PBR, shadow mapping, post-processing, bloom, SSAO, shader, instancing, LOD, WebXR, VR, AR, product configurator, data visualization, architectural walkthrough, interactive 3D, canvas. Use when: creating 3D visualizations, building WebGL/WebGPU apps, loading 3D models, adding animations, implementing VR/XR, creating interactive graphics, building product configurators.",{},"3D Graphics with Three.js","https://github.com/samhvw8/dot-claude/tree/HEAD/skills/3d-graphics",[317,318,319,320,321,16,322],"three-js","webgl","webgpu","3d-graphics","documentation","examples",{"githubOwner":324,"githubRepo":325,"locale":23,"slug":320,"type":190},"samhvw8","dot-claude",{"extract":327,"llm":329},{"commitSha":328,"license":196},"28c76162116d2eedab131c0e1548fdc76a2999f7",{"promptVersionExtension":168,"promptVersionScoring":169,"score":171,"targetMarket":174,"tier":175},{"repoId":331},"kd79ad9dpqazy79y2s6rvajgjn865xek",{"anyEnrichmentAt":333,"extractAt":334,"githubAt":333,"llmAt":335,"updatedAt":335},1778054813688,1778054812528,1778054896678,{"_creationTime":337,"_id":338,"community":339,"display":340,"identity":350,"providers":354,"relations":359,"workflow":361},1778053100136.2402,"k17e05w840mc0qc6pves9w7995866k14",{"reviewCount":8},{"description":341,"installMethods":342,"name":343,"sourceUrl":344,"tags":345},"Interact with DeerFlow AI agent platform via its HTTP API. Use this skill when the user wants to send messages or questions to DeerFlow for research/analysis, start a DeerFlow conversation thread, check DeerFlow status or health, list available models/skills/agents in DeerFlow, manage DeerFlow memory, upload files to DeerFlow threads, or delegate complex research tasks to DeerFlow. Also use when the user mentions deerflow, deer flow, or wants to run a deep research task that DeerFlow can handle.",{},"Claude to DeerFlow","https://github.com/bytedance/deer-flow/tree/HEAD/skills/public/claude-to-deerflow",[346,288,347,15,348,349],"deerflow","http","bash","python",{"githubOwner":351,"githubRepo":352,"locale":23,"slug":353,"type":190},"bytedance","deer-flow","claude-to-deerflow",{"extract":355,"llm":357},{"commitSha":356,"license":196},"1336872b15c25d45ebcb7c1cf72369c2bdd53187",{"promptVersionExtension":168,"promptVersionScoring":169,"score":358,"targetMarket":174,"tier":175},96,{"repoId":360},"kd789sm7egx1h0t1jag6zzhcq98656wv",{"anyEnrichmentAt":362,"extractAt":363,"githubAt":362,"llmAt":364,"updatedAt":364},1778053101076,1778053100136,1778053169012,{"_creationTime":366,"_id":367,"community":368,"display":369,"identity":379,"providers":383,"relations":388,"workflow":390},1778054086261.0896,"k173qc4w9k717e5jcb4da4zrcx866a1a",{"reviewCount":8},{"description":370,"name":371,"sourceUrl":372,"tags":373},"Recognize, diagnose, and mitigate patterns of context degradation in agent systems. Use when context grows large, agent performance degrades unexpectedly, or debugging agent failures.","Context Degradation Patterns","https://github.com/bilalmk/todo_correct/tree/HEAD/.claude/skills/mjs/context-degradation",[15,374,375,376,377,378],"context","performance","debugging","llm","analysis",{"githubOwner":380,"githubRepo":381,"locale":23,"slug":382,"type":190},"bilalmk","todo_correct","context-degradation",{"extract":384,"llm":386},{"commitSha":385,"license":196},"8b43aa04bd5c53e3cda46469b953684519a84ea7",{"promptVersionExtension":168,"promptVersionScoring":169,"score":387,"targetMarket":174,"tier":175},95,{"repoId":389},"kd75ecf652eb91ha327s8bqbex865z6v",{"anyEnrichmentAt":391,"extractAt":392,"githubAt":391,"llmAt":393,"updatedAt":393},1778054086910,1778054086261,1778054163453]