[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"extension-plugin-wshobson-plugin-eval-ru":3,"guides-for-wshobson-plugin-eval":264,"similar-k17fbgy6w2w7y8xhq8tzntb86d864vk2":265},{"_creationTime":4,"_id":5,"children":6,"community":7,"display":9,"evaluation":20,"identity":206,"isFallback":211,"parentExtension":212,"providers":246,"relations":250,"repo":251,"workflow":262},1778003443243.3257,"k17fbgy6w2w7y8xhq8tzntb86d864vk2",[],{"reviewCount":8},0,{"description":10,"installMethods":11,"name":12,"sourceUrl":13,"tags":14},"Three-layer quality evaluation framework for Claude Code plugins with Elo ranking",{},"PluginEval","https://github.com/wshobson/agents/tree/HEAD/plugins/plugin-eval",[15,16,17,18,19],"evaluation","quality","cli","testing","framework",{"_creationTime":21,"_id":22,"extensionId":5,"locale":23,"result":24,"trustSignals":195,"workflow":204},1778016883287.2312,"kn7f2ttwc4jtdjn8vy4gmqxpw5864pvq","en",{"checks":25,"evaluatedAt":185,"extensionSummary":186,"promptVersionExtension":187,"promptVersionScoring":188,"rationale":189,"score":190,"summary":191,"tags":192,"targetMarket":193,"tier":194},[26,31,35,38,42,46,50,54,57,60,64,69,72,77,80,83,86,89,92,95,98,102,106,110,114,117,120,123,127,130,133,136,139,142,145,147,149,152,154,156,158,161,164,167,171,174,177,181],{"category":27,"check":28,"severity":29,"summary":30},"Practical Utility","Problem relevance","info","The displayed description mentions \"Three-layer quality evaluation framework for Claude Code plugins with Elo ranking\", which is a good start but could be more specific about the problem it solves for users beyond just the technical framework.",{"category":27,"check":32,"severity":33,"summary":34},"Unique selling proposition","pass","The extension provides a unique, multi-layered quality evaluation framework with statistical backing (Elo ranking, Monte Carlo) that goes beyond simple static analysis, offering a comprehensive approach to assessing plugin quality.",{"category":27,"check":36,"severity":33,"summary":37},"Production readiness","The extension appears to be production-ready, offering a well-defined methodology, clear commands for evaluation (score, certify, compare), and a structured output format, indicating a mature and usable tool.",{"category":39,"check":40,"severity":33,"summary":41},"Scope","Single responsibility principle","The plugin-eval plugin focuses solely on evaluating the quality of other plugins and skills, adhering to a single, well-defined responsibility.",{"category":39,"check":43,"severity":44,"summary":45},"Description quality","warning","The displayed description is concise but lacks detail about the core problem it solves and the specific benefits of its methodology beyond the technical framework.",{"category":47,"check":48,"severity":33,"summary":49},"Invocation","Scoped tools","The plugin exposes specific commands like `/eval`, `/certify`, and `/compare`, which are narrowly scoped to their evaluation functions, rather than general-purpose shell execution.",{"category":51,"check":52,"severity":33,"summary":53},"Documentation","Configuration & parameter reference","The plugin's README and internal documentation (e.g., `docs/plugin-eval.md`) detail various commands, depths, and output formats, providing sufficient guidance for configuration and use.",{"category":39,"check":55,"severity":33,"summary":56},"Tool naming","The tools (`/eval`, `/certify`, `/compare`) are clearly named and descriptive of their functionality.",{"category":39,"check":58,"severity":33,"summary":59},"Minimal I/O surface","The CLI commands accept clear, specific arguments (path, depth, output format), and the output is structured (JSON, Markdown), minimizing ambiguity.",{"category":61,"check":62,"severity":33,"summary":63},"License","License usability","The extension includes an MIT License file, which is a permissive open-source license.",{"category":65,"check":66,"severity":67,"summary":68},"Maintenance","Commit recency","critical","The repository has no commit history (pushedAt: n/a), indicating it is likely unmaintained or has no active development, posing a significant risk for users.",{"category":65,"check":70,"severity":44,"summary":71},"Dependency Management","The pyproject.toml lists dependencies for dev and LLM integration, but there are no explicit measures like Dependabot configured or mentioned for managing or updating these dependencies, including potential vulnerabilities.",{"category":73,"check":74,"severity":75,"summary":76},"Security","Secret Management","not_applicable","The extension does not appear to handle any secrets directly.",{"category":73,"check":78,"severity":33,"summary":79},"Injection","The evaluation framework is designed to process specific file formats (SKILL.md, plugin.json) and commands, and the static analysis should treat input files as data, not instructions.",{"category":73,"check":81,"severity":33,"summary":82},"Transitive Supply-Chain Grenades","The extension appears to be self-contained and does not fetch external code or instructions at runtime, minimizing supply-chain risks.",{"category":73,"check":84,"severity":33,"summary":85},"Sandbox Isolation","The extension's operations are confined to processing local files for evaluation and do not involve modifying files outside the project's scope or interacting with the user's OS in a way that would break isolation.",{"category":73,"check":87,"severity":33,"summary":88},"Sandbox escape primitives","No evidence of detached processes or retry loops around denied calls was found in the provided scripts or documentation.",{"category":73,"check":90,"severity":33,"summary":91},"Data Exfiltration","The extension's primary function is local analysis, and it does not appear to exfiltrate any user data or credentials to third-party services.",{"category":73,"check":93,"severity":33,"summary":94},"Hidden Text Tricks","The bundled files (README, SKILL.md) do not contain any suspicious hidden text, Unicode characters, or ANSI escapes that could steer the model covertly.",{"category":73,"check":96,"severity":33,"summary":97},"Opaque code execution","The plugin's code appears to be plain Python, without obfuscation techniques like base64-encoded payloads or runtime script fetching.",{"category":99,"check":100,"severity":33,"summary":101},"Portability","Structural Assumption","The evaluation commands operate on provided paths and do not make assumptions about user-specific project layouts outside of these explicit inputs.",{"category":103,"check":104,"severity":67,"summary":105},"Trust","Issues Attention","No issues were opened or closed in the last 90 days (n/a), and the repository has no commit history, indicating a complete lack of maintainer engagement and slow responsiveness.",{"category":107,"check":108,"severity":44,"summary":109},"Versioning","Release Management","The project lacks a version number in its manifests (e.g., pyproject.toml, SKILL.md frontmatter), and there are no GitHub releases or CHANGELOG entries, making it difficult to track versions and rely on specific releases.",{"category":111,"check":112,"severity":33,"summary":113},"Code Execution","Validation","The CLI and internal logic appear to validate input paths and parameters to prevent basic errors, although detailed schema validation for all inputs is not explicitly clear.",{"category":73,"check":115,"severity":75,"summary":116},"Unguarded Destructive Operations","The extension is purely analytical and does not perform any destructive operations.",{"category":111,"check":118,"severity":33,"summary":119},"Error Handling","The code includes error handling for file operations and LLM queries, with mechanisms to return structured results or exceptions, which should allow the agent to handle errors gracefully.",{"category":111,"check":121,"severity":75,"summary":122},"Logging","The extension is a CLI tool for analysis and does not require runtime logging for audit purposes.",{"category":124,"check":125,"severity":75,"summary":126},"Compliance","GDPR","The extension performs local analysis and does not process personal data or interact with users in a way that would invoke GDPR concerns.",{"category":124,"check":128,"severity":33,"summary":129},"Target market","The extension is a code analysis tool and has no regional or jurisdictional logic, making it globally applicable.",{"category":99,"check":131,"severity":33,"summary":132},"Runtime stability","The extension relies on standard Python libraries and the `claude-agent-sdk`, which should ensure good cross-platform compatibility. It handles potential errors gracefully.",{"category":39,"check":134,"severity":75,"summary":135},"Tool surface size","This is a CLI tool and not a Claude Code plugin with exposed tools.",{"category":47,"check":137,"severity":75,"summary":138},"Name collisions","Not applicable as this is a CLI tool and not a Claude Code plugin being invoked within the Claude Code environment.",{"category":47,"check":140,"severity":75,"summary":141},"Overlapping near-synonym tools","Not applicable as this is a CLI tool and not a Claude Code plugin with exposed tools.",{"category":47,"check":143,"severity":75,"summary":144},"Hooks-off mechanism","Not applicable as this is a CLI tool and not a Claude Code plugin that utilizes hooks.",{"category":47,"check":146,"severity":75,"summary":144},"Hook matcher tightness",{"category":73,"check":148,"severity":75,"summary":144},"Hook security",{"category":150,"check":151,"severity":75,"summary":144},"Hooks","Silent prompt rewriting",{"category":73,"check":153,"severity":75,"summary":144},"Permission Hook",{"category":124,"check":155,"severity":75,"summary":144},"Hook privacy",{"category":111,"check":157,"severity":75,"summary":144},"Hook dependency",{"category":51,"check":159,"severity":33,"summary":160},"Install / Setup Instructions","The README provides clear setup instructions using `uv sync` and command examples for usage.",{"category":51,"check":162,"severity":33,"summary":163},"Feature Transparency","The README and internal documentation clearly outline the plugin's capabilities, commands, and evaluation layers.",{"category":51,"check":165,"severity":33,"summary":166},"Phantom features","All documented features (evaluation commands, layers) appear to have corresponding implementations in the codebase.",{"category":168,"check":169,"severity":33,"summary":170},"Convention","Layout convention adherence","The project follows standard Python packaging conventions with `pyproject.toml` and a `src/plugin_eval` layout.",{"category":168,"check":172,"severity":75,"summary":173},"Plugin state","Not applicable as this is a CLI tool and not a Claude Code plugin that manages persistent state.",{"category":73,"check":175,"severity":75,"summary":176},"Keychain-stored secrets","Not applicable as the tool does not handle secrets.",{"category":178,"check":179,"severity":33,"summary":180},"Dependencies","Tagged release sourcing","The plugin relies on published Python packages (e.g., pydantic, typer, claude-agent-sdk) managed via pyproject.toml, which are sourced from tagged releases.",{"category":182,"check":183,"severity":75,"summary":184},"Installation","Clean uninstall","Not applicable as this is a CLI tool installed via standard package management and does not spawn background processes.",1778016819668,"This extension provides a sophisticated three-layer quality evaluation framework for Claude Code plugins, utilizing static analysis, an LLM judge, and Monte Carlo simulations. It exposes CLI commands for scoring, certifying, and comparing plugins, offering detailed reports and anti-pattern detection.","2.0.0","3.4.0","The critical finding regarding the lack of commit recency and no issues attention, coupled with warnings about dependency management and release versioning, significantly impacts the trust and maintenance score. While the extension is functionally sound for its purpose, the absence of active maintenance and clear versioning makes it a risky choice for users.",30,"A comprehensive static and LLM-based quality evaluation framework for Claude Code plugins, but suffers from critical maintenance issues.",[15,16,17,18,19],"global","flagged",{"codeQuality":196,"collectedAt":197,"documentation":198,"maintenance":200,"security":201,"testCoverage":203},{},1778016804698,{"descriptionLength":199,"readmeSize":8},81,{},{"hasNpmPackage":202,"smitheryVerified":202},false,{"hasCi":202,"hasTests":202},{"updatedAt":205},1778016883287,{"githubOwner":207,"githubRepo":208,"locale":23,"slug":209,"type":210},"wshobson","agents","plugin-eval","plugin",true,{"_creationTime":213,"_id":214,"community":215,"display":216,"identity":228,"parentExtension":231,"providers":232,"relations":239,"workflow":241},1778003443243.2886,"k175qypm0s8m8k6a0fkxpxfj1n865ax9",{"reviewCount":8},{"description":217,"installMethods":218,"name":219,"sourceUrl":220,"tags":221},"Production-ready workflow orchestration with 79 focused plugins, 184 specialized agents, and 150 skills - optimized for granular installation and minimal token usage",{},"Claude Code Plugins: Orchestration and Automation","https://github.com/wshobson/agents",[222,223,224,225,226,227],"workflow-orchestration","ai-agents","plugins","development-tools","automation","developer-experience",{"githubOwner":207,"githubRepo":208,"locale":23,"slug":229,"type":230},"claude-code-workflows","marketplace",null,{"extract":233,"smithery":238},{"commitSha":234,"license":235,"marketplace":236},"ece811f23310a37ceb43496dbac0e244fe6845b6","MIT",{"name":229,"pluginCount":199,"version":237},"1.6.0",{"qualityScore":8,"totalActivations":8,"uniqueUsers":8,"useCount":8,"verified":202},{"repoId":240},"kd72tes1veaz04ac7p0d68ya4h8650m7",{"anyEnrichmentAt":242,"extractAt":243,"githubAt":244,"invalidatedAt":242,"llmAt":245,"smitheryAt":242,"updatedAt":242},1778016735335,1778003520097,1778003532786,1778016730286,{"extract":247,"llm":248,"smithery":249},{"commitSha":234,"license":235},{"promptVersionExtension":187,"promptVersionScoring":188,"score":190,"targetMarket":193,"tier":194},{"qualityScore":8,"totalActivations":8,"uniqueUsers":8,"useCount":8,"verified":202},{"parentExtensionId":214,"repoId":240},{"_creationTime":252,"_id":240,"identity":253,"providers":254,"workflow":260},1777995558409.8196,{"githubOwner":207,"githubRepo":208,"sourceUrl":220},{"discover":255},{"sources":256},[257,258,259],"skills-sh","smithery","vskill",{"discoverAt":261,"extractAt":243,"updatedAt":243},1777995558409,{"anyEnrichmentAt":263,"extractAt":243,"githubAt":244,"llmAt":205,"smitheryAt":263,"updatedAt":205},1778016861516,[],[266,302],{"_creationTime":267,"_id":268,"community":269,"display":270,"identity":282,"providers":286,"relations":294,"workflow":297},1778054452948.4272,"k179khyq4dvq0ytvdcepec984d8666wk",{"reviewCount":8},{"description":271,"name":272,"sourceUrl":273,"tags":274},"Comprehensive toolkit for developing Claude Code plugins. Includes 7 expert skills covering hooks, MCP integration, commands, agents, and best practices. AI-assisted plugin creation and validation.","Plugin Development Toolkit","https://github.com/anthropics/claude-plugins-official/tree/HEAD/plugins/plugin-dev",[275,276,277,208,278,279,17,226,280,281],"development","plugin-creation","hooks","skills","mcp","guidance","best-practices",{"githubOwner":283,"githubRepo":284,"locale":23,"slug":285,"type":210},"anthropics","claude-plugins-official","plugin-dev",{"extract":287,"llm":290,"smithery":293},{"commitSha":288,"license":289},"06f52cd3ac3e47ecb45228a86183ea2a86e9d6ff","Apache-2.0",{"promptVersionExtension":187,"promptVersionScoring":188,"score":291,"targetMarket":193,"tier":292},99,"verified",{"qualityScore":8,"totalActivations":8,"uniqueUsers":8,"useCount":8,"verified":202},{"parentExtensionId":295,"repoId":296},"k171b9714j6pgfxqht22y94q4x866sck","kd798hf3w99qz2xt1fqtgq7gf9865e31",{"anyEnrichmentAt":298,"extractAt":299,"githubAt":300,"llmAt":301,"smitheryAt":298,"updatedAt":301},1778054509977,1778054452948,1778054454391,1778054703946,{"_creationTime":303,"_id":304,"community":305,"display":306,"identity":317,"providers":319,"relations":324,"workflow":325},1778054452948.4097,"k17awzvhhaeaj8jcsnx8bcvgyd866nq9",{"reviewCount":8},{"description":307,"installMethods":308,"name":309,"sourceUrl":310,"tags":311},"Development kit for working with the Claude Agent SDK",{},"Agent SDK Development Plugin","https://github.com/anthropics/claude-plugins-official/tree/HEAD/plugins/agent-sdk-dev",[312,275,313,314,315,17,316],"sdk","typescript","python","agent","verification",{"githubOwner":283,"githubRepo":284,"locale":23,"slug":318,"type":210},"agent-sdk-dev",{"extract":320,"llm":321,"smithery":323},{"commitSha":288,"license":289},{"promptVersionExtension":187,"promptVersionScoring":188,"score":322,"targetMarket":193,"tier":292},98,{"qualityScore":8,"totalActivations":8,"uniqueUsers":8,"useCount":8,"verified":202},{"parentExtensionId":295,"repoId":296},{"anyEnrichmentAt":298,"extractAt":299,"githubAt":300,"llmAt":301,"smitheryAt":298,"updatedAt":301}]