[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"extension-skill-bilalmk-context-degradation-de":3,"guides-for-bilalmk-context-degradation":197,"similar-k173qc4w9k717e5jcb4da4zrcx866a1a":198},{"_creationTime":4,"_id":5,"children":6,"community":7,"display":9,"evaluation":20,"identity":164,"isFallback":169,"parentExtension":170,"providers":171,"relations":176,"repo":178,"workflow":194},1778054086261.0896,"k173qc4w9k717e5jcb4da4zrcx866a1a",[],{"reviewCount":8},0,{"description":10,"name":11,"sourceUrl":12,"tags":13},"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",[14,15,16,17,18,19],"agent","context","performance","debugging","llm","analysis",{"_creationTime":21,"_id":22,"extensionId":5,"locale":23,"result":24,"trustSignals":153,"workflow":162},1778054163453.6978,"kn7dpjm30b1pm1pqdgzhw1tnh98660f4","en",{"checks":25,"evaluatedAt":143,"extensionSummary":144,"promptVersionExtension":145,"promptVersionScoring":146,"rationale":147,"score":148,"summary":149,"tags":150,"targetMarket":151,"tier":152},[26,31,35,38,42,45,48,52,56,60,63,66,70,74,78,81,85,88,91,94,97,100,103,106,109,113,117,120,123,126,129,133,136,139,141],{"category":27,"check":28,"severity":29,"summary":30},"Invocation","Precise Purpose","pass","The description clearly states the skill's purpose: to recognize, diagnose, and mitigate context degradation patterns. It also provides specific triggers for when to use the skill, such as when agent performance degrades or when debugging failures.",{"category":32,"check":33,"severity":29,"summary":34},"Practical Utility","Unique selling proposition","The skill addresses a nuanced problem in agent systems (context degradation) that goes beyond basic prompt engineering, offering specific diagnostic and mitigation techniques rather than just being a thin wrapper around an API.",{"category":32,"check":36,"severity":29,"summary":37},"Production readiness","The skill provides detailed explanations, code examples, and architectural patterns for diagnosing and mitigating context degradation, suggesting it is ready for use in workflows that require handling large contexts.",{"category":39,"check":40,"severity":29,"summary":41},"Scope","Single responsibility principle","The skill focuses solely on the domain of context degradation in agent systems, providing tools and explanations related to this specific problem without venturing into unrelated areas like testing or deployment.",{"category":39,"check":43,"severity":29,"summary":44},"Description quality","The provided 'Displayed Description' accurately reflects the skill's capabilities and purpose as detailed in the SKILL.md file and its accompanying references.",{"category":27,"check":46,"severity":29,"summary":47},"Concise Frontmatter","The frontmatter in SKILL.md is concise, clearly defining the skill's name and purpose, and including specific triggers for invocation.",{"category":49,"check":50,"severity":29,"summary":51},"Documentation","Concise Body","The main body of the SKILL.md file is under 500 lines and effectively delegates detailed technical information to separate reference files, adhering to progressive disclosure principles.",{"category":53,"check":54,"severity":29,"summary":55},"Context","Progressive Disclosure","The SKILL.md outlines the overall flow and explicitly links to detailed technical references in the 'references' directory, demonstrating effective progressive disclosure.",{"category":53,"check":57,"severity":58,"summary":59},"Forked exploration","not_applicable","This skill is diagnostic and informational, not an exploration-heavy task that would necessitate forked exploration.",{"category":32,"check":61,"severity":29,"summary":62},"Usage examples","The skill includes two clear, ready-to-use examples demonstrating context degradation and mitigation strategies, with plausible outcomes.",{"category":32,"check":64,"severity":29,"summary":65},"Edge cases","The skill's instructions and references discuss various context degradation patterns (lost-in-middle, poisoning, distraction, confusion, clash) and provide mitigation strategies, effectively handling edge cases.",{"category":67,"check":68,"severity":58,"summary":69},"Code Execution","Tool Fallback","The skill does not appear to rely on external tools like an MCP server; it primarily uses Python scripts for analysis and documentation for guidance.",{"category":71,"check":72,"severity":29,"summary":73},"Portability","Stack assumptions","The included Python scripts are standard Python 3 and clearly state their assumptions in docstrings. The SKILL.md file provides general guidance without assuming specific tooling beyond standard LLM interaction.",{"category":75,"check":76,"severity":29,"summary":77},"Safety","Halt on unexpected state","The Python scripts and documentation imply a fail-closed behavior, where unexpected states would lead to reports or errors rather than proceeding silently. The analytical nature of the skill suggests it wouldn't proceed with destructive actions without clear state.",{"category":71,"check":79,"severity":29,"summary":80},"Cross-skill coupling","The skill focuses on context degradation analysis and mitigation and does not appear to rely on or implicitly couple with other specific skills.",{"category":82,"check":83,"severity":58,"summary":84},"Security","Secret Management","The skill's scripts are analytical and do not handle or expose any secrets.",{"category":82,"check":86,"severity":29,"summary":87},"Injection","The Python scripts primarily analyze provided context and do not load untrusted external data for execution or instructions.",{"category":82,"check":89,"severity":29,"summary":90},"Transitive Supply-Chain Grenades","The skill's scripts are self-contained and do not fetch external code or data at runtime. The provided references are bundled within the repository.",{"category":82,"check":92,"severity":29,"summary":93},"Sandbox Isolation","The provided Python scripts are analytical and do not perform any operations that would modify files or interact with the system outside of their intended scope.",{"category":82,"check":95,"severity":29,"summary":96},"Sandbox escape primitives","The Python scripts do not contain any detached process spawns or retry loops around denied tool calls.",{"category":82,"check":98,"severity":29,"summary":99},"Data Exfiltration","The skill's scripts are purely analytical and do not make any outbound network calls, especially not for submitting confidential data.",{"category":82,"check":101,"severity":29,"summary":102},"Hidden Text Tricks","The bundled files (SKILL.md, references, scripts) are free of hidden-steering tricks, control characters, or invisible Unicode sequences.",{"category":82,"check":104,"severity":29,"summary":105},"Opaque code execution","The Python scripts included in the bundle are plain, readable source code and do not contain obfuscation like base64 payloads or eval calls.",{"category":71,"check":107,"severity":29,"summary":108},"Structural Assumption","The skill's scripts do not make assumptions about user-specific project organization outside of the provided context. Any path handling is relative or clearly defined.",{"category":110,"check":111,"severity":58,"summary":112},"Trust","Issues Attention","No issues data available for evaluation.",{"category":114,"check":115,"severity":29,"summary":116},"Versioning","Release Management","The SKILL.md frontmatter includes a 'Version: 1.0.0' field, providing a clear version signal.",{"category":67,"check":118,"severity":29,"summary":119},"Validation","The Python scripts use basic Python structures and docstrings for clarity, and the skill's documentation outlines input handling and potential errors, implying a level of validation for its analytical tasks.",{"category":82,"check":121,"severity":29,"summary":122},"Unguarded Destructive Operations","The skill is purely analytical and diagnostic, containing no destructive operations.",{"category":67,"check":124,"severity":29,"summary":125},"Error Handling","The Python scripts include docstrings and error handling patterns (e.g., `try-except` blocks in simulated functions), and the documentation describes failure modes and recovery paths.",{"category":67,"check":127,"severity":58,"summary":128},"Logging","The skill's scripts are analytical and do not perform actions that require local audit logging.",{"category":130,"check":131,"severity":58,"summary":132},"Compliance","GDPR","The skill operates on context data for analysis and does not appear to handle personal data directly or submit it to third parties.",{"category":130,"check":134,"severity":29,"summary":135},"Target market","The skill's focus on general agent context degradation patterns and its use of generic examples suggest it is globally applicable. No regional-specific logic was detected.",{"category":71,"check":137,"severity":29,"summary":138},"Runtime stability","The skill's Python scripts use standard libraries and the SKILL.md assumes general LLM interaction, indicating good portability across different environments.",{"category":27,"check":50,"severity":29,"summary":140},"The skill body stays under ~500 lines and delegates deeper material to separate files (e.g., `references/patterns.md`), following progressive disclosure.",{"category":53,"check":54,"severity":29,"summary":142},"The SKILL.md effectively outlines the flow and links into `references/*.md` files for detailed technical information, demonstrating good progressive disclosure.",1778054125311,"This skill analyzes context length and content to identify and provide strategies for overcoming phenomena like the 'lost-in-middle' effect, context poisoning, distraction, confusion, and clash. It includes Python scripts for simulating attention analysis and poisoning detection, along with detailed markdown references on empirical benchmarks and recovery procedures.","2.0.0","3.4.0","The skill is excellent, with a clear purpose, robust documentation, and well-structured code. It excels in production readiness, documentation quality, and security. No critical or warning findings were found.",95,"A comprehensive skill for diagnosing and mitigating context degradation in agent systems.",[14,15,16,17,18,19],"global","verified",{"codeQuality":154,"collectedAt":155,"documentation":156,"maintenance":158,"security":159,"testCoverage":161},{},1778054115225,{"descriptionLength":157,"readmeSize":8},183,{},{"hasNpmPackage":160,"smitheryVerified":160},false,{"hasCi":160,"hasTests":160},{"updatedAt":163},1778054163453,{"githubOwner":165,"githubRepo":166,"locale":23,"slug":167,"type":168},"bilalmk","todo_correct","context-degradation","skill",true,null,{"extract":172,"llm":175},{"commitSha":173,"license":174},"8b43aa04bd5c53e3cda46469b953684519a84ea7","MIT",{"promptVersionExtension":145,"promptVersionScoring":146,"score":148,"targetMarket":151,"tier":152},{"repoId":177},"kd75ecf652eb91ha327s8bqbex865z6v",{"_creationTime":179,"_id":177,"identity":180,"providers":182,"workflow":191},1777995558409.9006,{"githubOwner":165,"githubRepo":166,"sourceUrl":181},"https://github.com/bilalmk/todo_correct",{"discover":183,"github":186},{"sources":184},[185],"skills-sh",{"closedIssues90d":8,"forks":8,"openIssues90d":187,"pushedAt":188,"readmeSize":189,"stars":187,"topics":190},1,1769509251000,14662,[],{"discoverAt":192,"extractAt":193,"githubAt":193,"updatedAt":193},1777995558409,1778054088050,{"anyEnrichmentAt":195,"extractAt":196,"githubAt":195,"llmAt":163,"updatedAt":163},1778054086910,1778054086261,[],[199,227,262,285,314,343],{"_creationTime":200,"_id":201,"community":202,"display":203,"identity":212,"providers":216,"relations":221,"workflow":223},1778053100136.2388,"k17ba7hx1c2htdr84qc7vc86cd867abn",{"reviewCount":8},{"description":204,"installMethods":205,"name":206,"sourceUrl":207,"tags":208},"Use this skill when the user requests to review, analyze, critique, or summarize academic papers, research articles, preprints, or scientific publications. Supports comprehensive structured reviews covering methodology assessment, contribution evaluation, literature positioning, and constructive feedback generation. Trigger on queries involving paper URLs, uploaded PDFs, arXiv links, or requests like \"review this paper\", \"analyze this research\", \"summarize this study\", or \"write a peer review\".",{},"Academic Paper Review Skill","https://github.com/bytedance/deer-flow/tree/HEAD/skills/public/academic-paper-review",[209,210,211,19,18],"research","academic","paper-review",{"githubOwner":213,"githubRepo":214,"locale":23,"slug":215,"type":168},"bytedance","deer-flow","academic-paper-review",{"extract":217,"llm":219},{"commitSha":218,"license":174},"1336872b15c25d45ebcb7c1cf72369c2bdd53187",{"promptVersionExtension":145,"promptVersionScoring":146,"score":220,"targetMarket":151,"tier":152},98,{"repoId":222},"kd789sm7egx1h0t1jag6zzhcq98656wv",{"anyEnrichmentAt":224,"extractAt":225,"githubAt":224,"llmAt":226,"updatedAt":226},1778053101076,1778053100136,1778053169012,{"_creationTime":228,"_id":229,"community":230,"display":231,"identity":246,"providers":250,"relations":255,"workflow":257},1778003499408.389,"k17139s2gtfqd4hex0j4przsqd864esx",{"reviewCount":8},{"description":232,"installMethods":233,"name":234,"sourceUrl":235,"tags":236},"Use when writing Playwright tests, fixing flaky tests, debugging failures, implementing Page Object Model, configuring CI/CD, optimizing performance, mocking APIs, handling authentication or OAuth, testing accessibility (axe-core), file uploads/downloads, date/time mocking, WebSockets, geolocation, permissions, multi-tab/popup flows, mobile/responsive layouts, touch gestures, GraphQL, error handling, offline mode, multi-user collaboration, third-party services (payments, email verification), console error monitoring, global setup/teardown, test annotations (skip, fixme, slow), test tags (@smoke, @fast, @critical, filtering with --grep), project dependencies, security testing (XSS, CSRF, auth), performance budgets (Web Vitals, Lighthouse), iframes, component testing, canvas/WebGL, service workers/PWA, test coverage, i18n/localization, Electron apps, or browser extension testing. Covers E2E, component, API, visual, accessibility, security, Electron, and extension testing.",{},"Playwright Best Practices Skill","https://github.com/currents-dev/playwright-best-practices-skill",[237,238,239,17,240,241,242,243,244,16,245],"playwright","testing","typescript","automation","e2e","component-testing","api-testing","visual-regression","security",{"githubOwner":247,"githubRepo":248,"locale":23,"slug":249,"type":168},"currents-dev","playwright-best-practices-skill","playwright-best-practices",{"extract":251,"llm":253,"smithery":254},{"commitSha":252,"license":174},"ef329e7e65149918e1ff0eed2cf7d2e6e6f9eb5b",{"promptVersionExtension":145,"promptVersionScoring":146,"score":148,"targetMarket":151,"tier":152},{"qualityScore":8,"totalActivations":8,"uniqueUsers":8,"useCount":8,"verified":160},{"repoId":256},"kd72s6aarbd4d9ag9h06119fw98640wa",{"anyEnrichmentAt":258,"extractAt":259,"githubAt":260,"llmAt":261,"smitheryAt":258,"updatedAt":261},1778016861516,1778003565910,1778003566667,1778016883287,{"_creationTime":263,"_id":264,"community":265,"display":266,"identity":275,"providers":277,"relations":283,"workflow":284},1778054086261.0906,"k173essf79wsmej31f7hydqnj5867s8q",{"reviewCount":8},{"description":267,"installMethods":268,"name":269,"sourceUrl":270,"tags":271},"Apply optimization techniques to extend effective context capacity. Use when context limits constrain agent performance, when optimizing for cost or latency, or when implementing long-running agent systems.",{},"Context Optimization Techniques","https://github.com/bilalmk/todo_correct/tree/HEAD/.claude/skills/mjs/context-optimization",[15,272,18,16,273,274],"optimization","cost","latency",{"githubOwner":165,"githubRepo":166,"locale":23,"slug":276,"type":168},"context-optimization",{"extract":278,"llm":280},{"commitSha":173,"license":279},"MIT-0",{"promptVersionExtension":145,"promptVersionScoring":146,"score":281,"targetMarket":151,"tier":282},49,"flagged",{"repoId":177},{"anyEnrichmentAt":195,"extractAt":196,"githubAt":195,"llmAt":163,"updatedAt":163},{"_creationTime":286,"_id":287,"community":288,"display":289,"identity":299,"providers":303,"relations":308,"workflow":310},1778053327521.585,"k17eftejnaapgr7mnfajqgng2h866syh",{"reviewCount":8},{"description":290,"installMethods":291,"name":292,"sourceUrl":293,"tags":294},"Systematic fact verification and misinformation identification using evidence-based analysis. Use when: verifying claims, checking facts, identifying misinformation, evaluating source credibility, or when user asks to \"fact check\", \"verify\", \"is this true\", or mentions claims that need validation.",{},"Fact Checker","https://github.com/shubhamsaboo/awesome-llm-apps/tree/HEAD/awesome_agent_skills/fact-checker",[295,296,297,298,19],"fact-checking","misinformation","verification","llm-skill",{"githubOwner":300,"githubRepo":301,"locale":23,"slug":302,"type":168},"shubhamsaboo","awesome-llm-apps","fact-checker",{"extract":304,"llm":306},{"commitSha":305,"license":174},"a35897449fe8b0fab12e8f0fd9f2e2a40e872ab7",{"promptVersionExtension":145,"promptVersionScoring":146,"score":307,"targetMarket":151,"tier":152},100,{"repoId":309},"kd73kvct1kme7748mpsbddhhmx865wd3",{"anyEnrichmentAt":311,"extractAt":312,"githubAt":311,"llmAt":313,"updatedAt":313},1778053329769,1778053327521,1778053376632,{"_creationTime":315,"_id":316,"community":317,"display":318,"identity":328,"providers":332,"relations":337,"workflow":339},1778053622473.6594,"k171jc4epkmqq2y76n42e3eg0s866rf8",{"reviewCount":8},{"description":319,"installMethods":320,"name":321,"sourceUrl":322,"tags":323},"Systematically identify what's missing in non-fiction writing—both blind spots (inherent limitations) and blank spots (gaps that could be addressed). Use before finalizing non-fiction or when feedback feels incomplete.",{},"Blind Spot Detective","https://github.com/jwynia/agent-skills/tree/HEAD/skills/general/writing/analysis/blind-spot-detective",[324,19,325,326,327],"writing","non-fiction","diagnostic","documentation",{"githubOwner":329,"githubRepo":330,"locale":23,"slug":331,"type":168},"jwynia","agent-skills","blind-spot-detective",{"extract":333,"llm":335},{"commitSha":334,"license":174},"e02ec7e226a6e4f8419fd3b88a1d8e472d421b32",{"promptVersionExtension":145,"promptVersionScoring":146,"score":336,"targetMarket":151,"tier":152},99,{"repoId":338},"kd7efn3mprpa8rd8vm5hw5ebzx864fph",{"anyEnrichmentAt":340,"extractAt":341,"githubAt":340,"llmAt":342,"updatedAt":342},1778053625386,1778053622473,1778054012696,{"_creationTime":344,"_id":345,"community":346,"display":347,"identity":357,"providers":360,"relations":364,"workflow":366},1778054617416.4724,"k171wbhkefsa1s7m82mat9azn9867e51",{"reviewCount":8},{"description":348,"installMethods":349,"name":350,"sourceUrl":351,"tags":352},"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",[14,353,354,355,356],"learning","self-improvement","logging","experience-capture",{"githubOwner":358,"githubRepo":359,"locale":23,"slug":359,"type":168},"yanhongxi-openclaw","proactive-self-improving-agent",{"extract":361,"llm":363},{"commitSha":362,"license":174},"5af957709186392fb9892382d0ffb26e8095ad65",{"promptVersionExtension":145,"promptVersionScoring":146,"score":220,"targetMarket":151,"tier":152},{"repoId":365},"kd78yyb7ypqjg4av4r3p0feyfd864sdm",{"anyEnrichmentAt":367,"extractAt":368,"githubAt":367,"llmAt":369,"updatedAt":369},1778054617695,1778054617416,1778054631760]