[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"extension-skill-Whatsonyourmind-oraclaw-bandit-en":3,"guides-for-Whatsonyourmind-oraclaw-bandit":433,"similar-k17ewnzxwqzf8m1b1t3c7srke186nx9a-en":434},{"_creationTime":4,"_id":5,"children":6,"community":7,"display":9,"evaluation":15,"identity":253,"isFallback":248,"parentExtension":259,"providers":260,"relations":265,"repo":267,"tags":429,"workflow":430},1778698837670.7983,"k17ewnzxwqzf8m1b1t3c7srke186nx9a",[],{"reviewCount":8},0,{"description":10,"installMethods":11,"name":13,"sourceUrl":14},"A/B testing and feature optimization for AI agents. Pick the best option automatically using Multi-Armed Bandits and Contextual Bandits (LinUCB). No data warehouse needed — works from request",{"claudeCode":12},"Whatsonyourmind/oraclaw","OraClaw Bandit","https://github.com/Whatsonyourmind/oraclaw",{"_creationTime":16,"_id":17,"extensionId":5,"locale":18,"result":19,"trustSignals":234,"workflow":251},1778698869396.3748,"kn7egapdf3xrtnbkyj5ccakbk986nfrc","en",{"checks":20,"evaluatedAt":195,"extensionSummary":196,"features":197,"nonGoals":203,"practices":207,"prerequisites":211,"promptVersionExtension":214,"promptVersionScoring":215,"purpose":216,"rationale":217,"score":218,"summary":219,"tags":220,"targetMarket":227,"tier":228,"useCases":229},[21,26,29,33,37,41,45,48,52,56,60,63,66,69,72,76,79,82,85,88,91,95,99,102,106,109,112,115,118,121,124,128,131,135,139,142,145,148,151,155,158,161,164,167,170,173,177,181,185,188,192],{"category":22,"check":23,"severity":24,"summary":25},"Invocation","Precise Purpose","pass","The description clearly states the skill's purpose (A/B testing and feature optimization using bandits) and when to use it, naming the artifact (options/variants) and user intent (choose the best option, run tests, optimize features).",{"category":22,"check":27,"severity":24,"summary":28},"Concise Frontmatter","The frontmatter is concise, self-contained, and clearly summarizes the core capability and use cases within the first ~160 characters.",{"category":30,"check":31,"severity":24,"summary":32},"Documentation","Concise Body","The SKILL.md is concise and delegates deeper material to separate files, keeping the main body under reasonable length.",{"category":34,"check":35,"severity":24,"summary":36},"Context","Progressive Disclosure","The SKILL.md outlines the flow and references external details where appropriate, avoiding excessive inline data.",{"category":34,"check":38,"severity":39,"summary":40},"Forked exploration","not_applicable","This skill is not designed for deep exploration or research that would flood the conversation, thus 'context: fork' is not applicable.",{"category":42,"check":43,"severity":24,"summary":44},"Practical Utility","Usage examples","Sufficient end-to-end examples are provided for major capabilities, demonstrating input, invocation, and observable outcomes, and they appear plausible.",{"category":42,"check":46,"severity":24,"summary":47},"Edge cases","The skill handles edge cases and limitations, documenting failure modes with observable symptoms and recovery steps.",{"category":49,"check":50,"severity":24,"summary":51},"Code Execution","Tool Fallback","The skill lists the required MCP and notes it is optional with a fallback, aligning with documentation best practices.",{"category":53,"check":54,"severity":24,"summary":55},"Portability","Stack assumptions","The skill clearly states its stack assumptions, including the runtime environment (Node.js via npx) and minimum versions where applicable, and dependencies are declared.",{"category":57,"check":58,"severity":24,"summary":59},"Safety","Halt on unexpected state","The skill lists preconditions and implies halting on unexpected state, ensuring a safe workflow.",{"category":53,"check":61,"severity":24,"summary":62},"Cross-skill coupling","The skill is self-contained and does not implicitly rely on other skills, with explicit cross-linking for related functionality.",{"category":42,"check":64,"severity":24,"summary":65},"Problem relevance","The description clearly names the user problem of needing AI-driven optimization for choices and A/B testing without complex data warehousing.",{"category":42,"check":67,"severity":24,"summary":68},"Unique selling proposition","The extension offers a unique selling proposition by providing deterministic mathematical optimization algorithms (bandits, LinUCB) for AI agents, going beyond simple prompting and offering low-latency, token-free computation.",{"category":42,"check":70,"severity":24,"summary":71},"Production readiness","The extension is production-ready, offering multiple integration methods (MCP server, REST API, SDK) and covering the complete lifecycle of optimization tasks.",{"category":73,"check":74,"severity":24,"summary":75},"Scope","Single responsibility principle","The extension focuses on optimization tasks using bandit algorithms, adhering to a single, well-defined responsibility.",{"category":73,"check":77,"severity":24,"summary":78},"Description quality","The description is accurate, concise, readable, and accurately reflects the skill's behavior and capabilities.",{"category":22,"check":80,"severity":24,"summary":81},"Scoped tools","The extension exposes narrow verb-noun specialist tools like `optimize_bandit` and `optimize_contextual`.",{"category":30,"check":83,"severity":24,"summary":84},"Configuration & parameter reference","All parameters for the bandit and contextual tools are documented, including context vector consistency and reward normalization.",{"category":73,"check":86,"severity":24,"summary":87},"Tool naming","Tool names like `optimize_bandit` and `optimize_contextual` are descriptive and follow a clear convention.",{"category":73,"check":89,"severity":24,"summary":90},"Minimal I/O surface","Tool inputs are structured (JSON objects with defined fields) and outputs are minimal, returning only the selected arm and relevant scores.",{"category":92,"check":93,"severity":24,"summary":94},"License","License usability","The extension is distributed under the MIT license, which is permissive and clearly stated in both the README and LICENSE file.",{"category":96,"check":97,"severity":24,"summary":98},"Maintenance","Commit recency","The latest commit was on May 2nd, 2026, well within the last 90 days, indicating active maintenance.",{"category":96,"check":100,"severity":24,"summary":101},"Dependency Management","The project utilizes npm and clearly lists dependencies, with the `mcp-server` likely managed via npm, indicating good dependency management practices.",{"category":103,"check":104,"severity":24,"summary":105},"Security","Secret Management","Secrets like `ORACLAW_API_KEY` are handled via environment variables and are not echoed into output, with appropriate measures in place.",{"category":103,"check":107,"severity":24,"summary":108},"Injection","The extension does not appear to load or execute untrusted third-party data as instructions, and does not fetch remote content as executable code.",{"category":103,"check":110,"severity":24,"summary":111},"Transitive Supply-Chain Grenades","The extension does not fetch external files at runtime or use remote-pipe-to-shell patterns; all dependencies are managed and committed.",{"category":103,"check":113,"severity":24,"summary":114},"Sandbox Isolation","The extension operates within the defined project folder and does not modify files outside of its scope, respecting sandbox isolation.",{"category":103,"check":116,"severity":24,"summary":117},"Sandbox escape primitives","No detached-process spawns or deny-retry loops were found in the provided scripts, indicating no sandbox escape primitives.",{"category":103,"check":119,"severity":24,"summary":120},"Data Exfiltration","The extension does not instruct the agent to read or submit confidential data to third parties. Outbound calls are documented and for known services.",{"category":103,"check":122,"severity":24,"summary":123},"Hidden Text Tricks","Bundled content appears free of hidden-steering tricks, and descriptions use clean printable ASCII.",{"category":125,"check":126,"severity":24,"summary":127},"Hooks","Opaque code execution","The provided scripts and code are plain and readable, with no obfuscation like base64 payloads or runtime fetching of code.",{"category":53,"check":129,"severity":24,"summary":130},"Structural Assumption","The skill does not make structural assumptions about user-specific project organization outside of its bundle and clearly states its requirements.",{"category":132,"check":133,"severity":24,"summary":134},"Trust","Issues Attention","With 0 issues opened and 44 closed in the last 90 days, the closure rate is high and maintainer engagement is strong.",{"category":136,"check":137,"severity":24,"summary":138},"Versioning","Release Management","The extension has a clear version (`1.0.0`) in the SKILL.md frontmatter and CHANGELOG.md, indicating proper release management.",{"category":49,"check":140,"severity":24,"summary":141},"Validation","Input arguments for the tools are clearly defined in schemas and validated, ensuring robust execution.",{"category":103,"check":143,"severity":24,"summary":144},"Unguarded Destructive Operations","This extension is read-only in nature and does not perform destructive operations.",{"category":49,"check":146,"severity":24,"summary":147},"Error Handling","Errors are caught and reported meaningfully, with structured fields provided for agent decision-making, and no silent retries of destructive actions.",{"category":49,"check":149,"severity":39,"summary":150},"Logging","The skill does not perform destructive actions or outbound calls that require local audit logging.",{"category":152,"check":153,"severity":39,"summary":154},"Compliance","GDPR","The extension does not operate on personal data and therefore does not require specific GDPR sanitization measures.",{"category":152,"check":156,"severity":24,"summary":157},"Target market","The extension's algorithms are globally applicable, and no regional or jurisdictional restrictions are indicated.",{"category":53,"check":159,"severity":24,"summary":160},"Runtime stability","The extension is designed to run on standard environments (Node.js via npx) and does not make assumptions about specific editors or OS-specific features.",{"category":30,"check":162,"severity":24,"summary":163},"README","The README exists, is comprehensive, and clearly states the extension's purpose and capabilities.",{"category":73,"check":165,"severity":24,"summary":166},"Tool surface size","The extension exposes 17 tools, which is within the desired range for a comprehensive MCP server.",{"category":22,"check":168,"severity":24,"summary":169},"Overlapping near-synonym tools","Tool names are distinct and cover specific optimization tasks without significant overlap or near-synonym redundancy.",{"category":30,"check":171,"severity":24,"summary":172},"Phantom features","Every feature advertised in the README and SKILL.md has a corresponding implementation in the provided tools.",{"category":174,"check":175,"severity":24,"summary":176},"Install","Installation instruction","Clear installation instructions are provided for MCP server setup, REST API usage, and npm SDK, with copy-paste examples.",{"category":178,"check":179,"severity":24,"summary":180},"Errors","Actionable error messages","Error paths are well-defined, with clear messages that include what failed, why, and remediation steps or links.",{"category":182,"check":183,"severity":24,"summary":184},"Execution","Pinned dependencies","Dependencies are managed via npm and the MCP server setup implies pinned versions, ensuring reproducible builds.",{"category":73,"check":186,"severity":39,"summary":187},"Dry-run preview","The extension is primarily analytical and does not perform state-changing operations that would require a dry-run preview.",{"category":189,"check":190,"severity":24,"summary":191},"Protocol","Idempotent retry & timeouts","The extension's operations are idempotent where possible, and the MCP server and API are designed for statelessness with timeouts.",{"category":152,"check":193,"severity":24,"summary":194},"Telemetry opt-in","Telemetry is not mentioned as being collected, implying no opt-out telemetry, and any potential collection would be opt-in based on standard practices.",1778698869231,"This skill provides AI agents with advanced optimization capabilities using Multi-Armed Bandits and Contextual Bandits (LinUCB). It offers deterministic mathematical solutions for choosing between options, running A/B tests, and personalizing selections based on context, without requiring a data warehouse.",[198,199,200,201,202],"Automatic selection of best variants using bandits","Context-aware optimization with LinUCB","Low-latency (\u003C25ms) and token-free computations","Multiple integration methods (MCP server, REST API, SDK)","Support for various optimization algorithms",[204,205,206],"Performing arbitrary mathematical calculations beyond optimization","Acting as a general-purpose data analysis or warehousing tool","Replacing LLM reasoning for tasks that do not require deterministic mathematical solutions",[208,209,210],"Optimization","Experimentation Design","Machine Learning Operations",[212,213],"ORACLAW_API_KEY environment variable for premium features","Node.js/npm for local MCP server setup","3.0.0","4.4.0","To equip AI agents with precise, deterministic optimization algorithms for decision-making, enabling them to select the best options, run effective A/B tests, and optimize features without relying on potentially fallible LLM heuristics.","High quality across all checks, particularly strong in practical utility, documentation, and security. The only notable area is the lack of a dry-run feature, which is not applicable given the extension's analytical nature.",99,"A robust and production-ready skill for AI-powered A/B testing and optimization.",[221,222,223,224,225,226],"optimization","ab-testing","feature-flags","experimentation","machine-learning","mcp","global","verified",[230,231,232,233],"Choosing the best variant from multiple options for A/B tests","Optimizing feature flags, prompts, email subjects, or any choice","Making context-aware selections based on user, time, or situation","Running adaptive experiments without predetermined sample sizes",{"codeQuality":235,"collectedAt":237,"documentation":238,"maintenance":241,"security":247,"testCoverage":250},{"hasLockfile":236},true,1778698853904,{"descriptionLength":239,"readmeSize":240},191,9472,{"closedIssues90d":242,"forks":243,"hasChangelog":236,"manifestVersion":244,"openIssues90d":8,"pushedAt":245,"stars":246},44,2,"1.0.0",1777714123000,8,{"hasNpmPackage":248,"license":249,"smitheryVerified":248},false,"MIT",{"hasCi":236,"hasTests":236},{"updatedAt":252},1778698869396,{"basePath":254,"githubOwner":255,"githubRepo":256,"locale":18,"slug":257,"type":258},"mission-control/packages/clawhub-skills/oraclaw-bandit","Whatsonyourmind","oraclaw","oraclaw-bandit","skill",null,{"evaluate":261,"extract":263},{"promptVersionExtension":214,"promptVersionScoring":215,"score":218,"tags":262,"targetMarket":227,"tier":228},[221,222,223,224,225,226],{"commitSha":264,"license":249},"HEAD",{"repoId":266},"kd76fmxm1ng903s4fmj0p7hxxs86ndkg",{"_creationTime":268,"_id":266,"identity":269,"providers":270,"workflow":425},1778698831609.0093,{"githubOwner":255,"githubRepo":256,"sourceUrl":14},{"classify":271,"discover":401,"github":404},{"commitSha":264,"extensions":272},[273,284,289,297,305,313,321,329,337,345,353,361,369,377,385],{"basePath":274,"description":275,"displayName":276,"installMethods":277,"rationale":278,"selectedPaths":279,"source":283,"sourceLanguage":18,"type":258},"mission-control/packages/clawhub-skills/oraclaw-anomaly","Anomaly detection for AI agents. Z-score, IQR, and streaming detection. Find outliers in data instantly. Sub-millisecond response. Works on single values or full datasets.","oraclaw-anomaly",{"claudeCode":12},"SKILL.md frontmatter at mission-control/packages/clawhub-skills/oraclaw-anomaly/SKILL.md",[280],{"path":281,"priority":282},"SKILL.md","mandatory","rule",{"basePath":254,"description":10,"displayName":257,"installMethods":285,"rationale":286,"selectedPaths":287,"source":283,"sourceLanguage":18,"type":258},{"claudeCode":12},"SKILL.md frontmatter at mission-control/packages/clawhub-skills/oraclaw-bandit/SKILL.md",[288],{"path":281,"priority":282},{"basePath":290,"description":291,"displayName":292,"installMethods":293,"rationale":294,"selectedPaths":295,"source":283,"sourceLanguage":18,"type":258},"mission-control/packages/clawhub-skills/oraclaw-bayesian","Bayesian inference engine for AI agents. Update beliefs with new evidence. Prior + evidence = posterior. Multi-factor prediction with calibration tracking.","oraclaw-bayesian",{"claudeCode":12},"SKILL.md frontmatter at mission-control/packages/clawhub-skills/oraclaw-bayesian/SKILL.md",[296],{"path":281,"priority":282},{"basePath":298,"description":299,"displayName":300,"installMethods":301,"rationale":302,"selectedPaths":303,"source":283,"sourceLanguage":18,"type":258},"mission-control/packages/clawhub-skills/oraclaw-calibrate","Prediction quality scoring for AI agents. Brier score, log score, and multi-source convergence analysis. Know if your forecasts are accurate and if your data sources agree.","oraclaw-calibrate",{"claudeCode":12},"SKILL.md frontmatter at mission-control/packages/clawhub-skills/oraclaw-calibrate/SKILL.md",[304],{"path":281,"priority":282},{"basePath":306,"description":307,"displayName":308,"installMethods":309,"rationale":310,"selectedPaths":311,"source":283,"sourceLanguage":18,"type":258},"mission-control/packages/clawhub-skills/oraclaw-cmaes","CMA-ES continuous optimization for AI agents. State-of-the-art derivative-free optimizer. 10-100x more sample-efficient than genetic algorithms on continuous problems. Hyperparameter tuning, portfolio optimization, parameter calibration.","oraclaw-cmaes",{"claudeCode":12},"SKILL.md frontmatter at mission-control/packages/clawhub-skills/oraclaw-cmaes/SKILL.md",[312],{"path":281,"priority":282},{"basePath":314,"description":315,"displayName":316,"installMethods":317,"rationale":318,"selectedPaths":319,"source":283,"sourceLanguage":18,"type":258},"mission-control/packages/clawhub-skills/oraclaw-decide","Decision intelligence for AI agents. Analyze options, map decision dependencies with PageRank, detect when information sources conflict, and find the choices that matter most.","oraclaw-decide",{"claudeCode":12},"SKILL.md frontmatter at mission-control/packages/clawhub-skills/oraclaw-decide/SKILL.md",[320],{"path":281,"priority":282},{"basePath":322,"description":323,"displayName":324,"installMethods":325,"rationale":326,"selectedPaths":327,"source":283,"sourceLanguage":18,"type":258},"mission-control/packages/clawhub-skills/oraclaw-ensemble","Multi-model consensus for AI agents. Combine predictions from multiple LLMs, models, or sources into a mathematically optimal consensus. Auto-weights by historical accuracy.","oraclaw-ensemble",{"claudeCode":12},"SKILL.md frontmatter at mission-control/packages/clawhub-skills/oraclaw-ensemble/SKILL.md",[328],{"path":281,"priority":282},{"basePath":330,"description":331,"displayName":332,"installMethods":333,"rationale":334,"selectedPaths":335,"source":283,"sourceLanguage":18,"type":258},"mission-control/packages/clawhub-skills/oraclaw-evolve","Genetic Algorithm optimizer for AI agents. Multi-objective Pareto optimization for portfolio weights, pricing, hyperparameters, marketing mix — any problem with multiple competing goals. Handles nonlinear search spaces that LP solvers cannot.","oraclaw-evolve",{"claudeCode":12},"SKILL.md frontmatter at mission-control/packages/clawhub-skills/oraclaw-evolve/SKILL.md",[336],{"path":281,"priority":282},{"basePath":338,"description":339,"displayName":340,"installMethods":341,"rationale":342,"selectedPaths":343,"source":283,"sourceLanguage":18,"type":258},"mission-control/packages/clawhub-skills/oraclaw-forecast","Time series forecasting for AI agents. ARIMA and Holt-Winters predictions with confidence intervals. Predict revenue, traffic, prices, or any sequential data. Sub-5ms inference.","oraclaw-forecast",{"claudeCode":12},"SKILL.md frontmatter at mission-control/packages/clawhub-skills/oraclaw-forecast/SKILL.md",[344],{"path":281,"priority":282},{"basePath":346,"description":347,"displayName":348,"installMethods":349,"rationale":350,"selectedPaths":351,"source":283,"sourceLanguage":18,"type":258},"mission-control/packages/clawhub-skills/oraclaw-graph","Network intelligence for AI agents. PageRank, community detection (Louvain), critical path, and bottleneck analysis for any graph of connected things.","oraclaw-graph",{"claudeCode":12},"SKILL.md frontmatter at mission-control/packages/clawhub-skills/oraclaw-graph/SKILL.md",[352],{"path":281,"priority":282},{"basePath":354,"description":355,"displayName":356,"installMethods":357,"rationale":358,"selectedPaths":359,"source":283,"sourceLanguage":18,"type":258},"mission-control/packages/clawhub-skills/oraclaw-pathfind","A* pathfinding and task sequencing for AI agents. Find the optimal path through workflows, dependencies, and decision trees. K-shortest paths via Yen's algorithm. Cost/time/risk breakdown.","oraclaw-pathfind",{"claudeCode":12},"SKILL.md frontmatter at mission-control/packages/clawhub-skills/oraclaw-pathfind/SKILL.md",[360],{"path":281,"priority":282},{"basePath":362,"description":363,"displayName":364,"installMethods":365,"rationale":366,"selectedPaths":367,"source":283,"sourceLanguage":18,"type":258},"mission-control/packages/clawhub-skills/oraclaw-risk","Risk assessment engine for AI agents. Value at Risk (VaR), CVaR, stress testing, and multi-factor risk scoring. Monte Carlo powered. Built for trading agents, lending agents, and portfolio managers.","oraclaw-risk",{"claudeCode":12},"SKILL.md frontmatter at mission-control/packages/clawhub-skills/oraclaw-risk/SKILL.md",[368],{"path":281,"priority":282},{"basePath":370,"description":371,"displayName":372,"installMethods":373,"rationale":374,"selectedPaths":375,"source":283,"sourceLanguage":18,"type":258},"mission-control/packages/clawhub-skills/oraclaw-simulate","Monte Carlo simulation for AI agents. Run thousands of probabilistic scenarios to model risk, forecast revenue, estimate project timelines, and quantify uncertainty. Supports 6 distribution types.","oraclaw-simulate",{"claudeCode":12},"SKILL.md frontmatter at mission-control/packages/clawhub-skills/oraclaw-simulate/SKILL.md",[376],{"path":281,"priority":282},{"basePath":378,"description":379,"displayName":380,"installMethods":381,"rationale":382,"selectedPaths":383,"source":283,"sourceLanguage":18,"type":258},"mission-control/packages/clawhub-skills/oraclaw-solver","Industrial-grade scheduling and resource optimization for AI agents. Solve task scheduling with energy matching, budget allocation, and any LP/MIP constraint problem in milliseconds.","oraclaw-solver",{"claudeCode":12},"SKILL.md frontmatter at mission-control/packages/clawhub-skills/oraclaw-solver/SKILL.md",[384],{"path":281,"priority":282},{"basePath":386,"description":387,"displayName":388,"installMethods":389,"license":249,"rationale":390,"selectedPaths":391,"source":283,"sourceLanguage":18,"type":226},"mission-control/packages/mcp-server","OraClaw Decision Intelligence — 17 MCP tools for AI agents (6 premium API-key tools + 11 free). Full input/output schemas + MCP behavior annotations on every tool. Optimization (bandit/CMA-ES/genetic/LP-MIP), simulation (Monte Carlo/scenarios), prediction (ARIMA/Holt-Winters/Bayesian/ensemble), scoring (convergence/calibration), graph analytics, anomaly detection, pathfinding, scheduling.","@oraclaw/mcp-server",{"npm":388},"server.json with namespace/server name at mission-control/packages/mcp-server/server.json",[392,394,396,398],{"path":393,"priority":282},"server.json",{"path":395,"priority":282},"package.json",{"path":397,"priority":282},"README.md",{"path":399,"priority":400},"src/index.ts","low",{"sources":402},[403],"manual",{"closedIssues90d":242,"description":405,"forks":243,"homepage":406,"license":249,"openIssues90d":8,"pushedAt":245,"readmeSize":240,"stars":246,"topics":407},"Deterministic decision-intelligence MCP server for AI agents — 17 tools, 21 algorithms (LinUCB, HiGHS LP/MIP, PageRank, Monte Carlo, CMA-ES, conformal). Sub-25ms. Zero LLM cost. AAA on Glama. Field-validated in 12+ OSS projects.","https://web-olive-one-89.vercel.app",[408,409,410,411,412,413,225,226,221,414,415,416,417,418,419,420,421,422,423,424],"ai-agents","algorithms","api","bandits","decision-intelligence","fastify","typescript","agent-tools","anthropic","claude-mcp","contextual-bandit","deterministic-tools","linear-programming","llm-tools","model-context-protocol","monte-carlo","pagerank",{"classifiedAt":426,"discoverAt":427,"extractAt":428,"githubAt":428,"updatedAt":426},1778698837409,1778698831609,1778698835357,[222,224,223,225,226,221],{"evaluatedAt":252,"extractAt":431,"updatedAt":432},1778698837670,1778699186606,[],[435,464,493,520,542,570],{"_creationTime":436,"_id":437,"community":438,"display":439,"identity":445,"providers":449,"relations":457,"tags":460,"workflow":461},1778695720086.7703,"k176r34g5a5fjn1z1a4gq6v88186nje0",{"reviewCount":8},{"description":440,"installMethods":441,"name":443,"sourceUrl":444},"Designs an A/B test or experiment with clear hypothesis, variants, success metrics, sample size, and duration. Use when planning experiments to validate product changes or test hypotheses.",{"claudeCode":442},"product-on-purpose/pm-skills","measure-experiment-design","https://github.com/product-on-purpose/pm-skills",{"basePath":446,"githubOwner":447,"githubRepo":448,"locale":18,"slug":443,"type":258},"skills/measure-experiment-design","product-on-purpose","pm-skills",{"evaluate":450,"extract":456},{"promptVersionExtension":214,"promptVersionScoring":215,"score":451,"tags":452,"targetMarket":227,"tier":228},100,[222,224,453,454,455],"product-management","a-b-testing","data-analysis",{"commitSha":264},{"parentExtensionId":458,"repoId":459},"k1721116hsfj7zg78w03432n8986n6y8","kd78ksv1wjj826ds5j1sh2kqnx86mhqf",[454,222,455,224,453],{"evaluatedAt":462,"extractAt":463,"updatedAt":462},1778696438706,1778695720086,{"_creationTime":465,"_id":466,"community":467,"display":468,"identity":474,"providers":479,"relations":486,"tags":489,"workflow":490},1778686640222.7952,"k178bs4zybvyebq2gnym4jazch86np03",{"reviewCount":8},{"description":469,"installMethods":470,"name":472,"sourceUrl":473},"Run metric-driven iterative optimization loops -- define a measurable goal, run parallel experiments, measure each against hard gates or LLM-as-judge scores, keep improvements, and converge on the best solution. Use when optimizing clustering quality, search relevance, build performance, prompt quality, or any measurable outcome that benefits from systematic experimentation.",{"claudeCode":471},"EveryInc/compound-engineering-plugin","CE Optimize","https://github.com/EveryInc/compound-engineering-plugin",{"basePath":475,"githubOwner":476,"githubRepo":477,"locale":18,"slug":478,"type":258},"plugins/compound-engineering/skills/ce-optimize","EveryInc","compound-engineering-plugin","ce-optimize",{"evaluate":480,"extract":485},{"promptVersionExtension":214,"promptVersionScoring":215,"score":451,"tags":481,"targetMarket":227,"tier":228},[221,224,482,483,484],"mlops","code-quality","prompt-engineering",{"commitSha":264,"license":249},{"parentExtensionId":487,"repoId":488},"k17d893df4em0e3pn8f55h1dxn86n09e","kd7e40my1b5g70tg0f60qg85ch86nn08",[483,224,482,221,484],{"evaluatedAt":491,"extractAt":492,"updatedAt":491},1778687141592,1778686640222,{"_creationTime":494,"_id":495,"community":496,"display":497,"identity":503,"providers":507,"relations":513,"tags":516,"workflow":517},1778675056600.2664,"k173ggr45da0wr1c32bva99bvh86nj5a",{"reviewCount":8},{"description":498,"installMethods":499,"name":501,"sourceUrl":502},"Use when planning product experiments, writing testable hypotheses, estimating sample size, prioritizing tests, or interpreting A/B outcomes with practical statistical rigor.",{"claudeCode":500},"alirezarezvani/claude-skills","experiment-designer","https://github.com/alirezarezvani/claude-skills",{"basePath":504,"githubOwner":505,"githubRepo":506,"locale":18,"slug":501,"type":258},"product-team/skills/experiment-designer","alirezarezvani","claude-skills",{"evaluate":508,"extract":512},{"promptVersionExtension":214,"promptVersionScoring":215,"score":218,"tags":509,"targetMarket":227,"tier":228},[224,222,453,510,455,511],"statistics","hypothesis-testing",{"commitSha":264},{"parentExtensionId":514,"repoId":515},"k17104ysr0smp7vvp26mnn1fzh86nynm","kd7ff9s1w43mfyy1n7hf87816186m6px",[222,455,224,511,453,510],{"evaluatedAt":518,"extractAt":519,"updatedAt":518},1778685452060,1778675056600,{"_creationTime":521,"_id":522,"community":523,"display":524,"identity":528,"providers":530,"relations":537,"tags":539,"workflow":540},1778675056600.2542,"k17d6x2qxdeyeegtqjtkz3myqx86mq6m",{"reviewCount":8},{"description":525,"installMethods":526,"name":527,"sourceUrl":502},"When the user wants to plan, design, or implement an A/B test or experiment. Also use when the user mentions \"A/B test,\" \"split test,\" \"experiment,\" \"test this change,\" \"variant copy,\" \"multivariate test,\" \"hypothesis,\" \"conversion experiment,\" \"statistical significance,\" or \"test this.\" For tracking implementation, see analytics-tracking.",{"claudeCode":500},"ab-test-setup",{"basePath":529,"githubOwner":505,"githubRepo":506,"locale":18,"slug":527,"type":258},"marketing-skill/skills/ab-test-setup",{"evaluate":531,"extract":536},{"promptVersionExtension":214,"promptVersionScoring":215,"score":532,"tags":533,"targetMarket":227,"tier":228},98,[222,224,534,535,455],"conversion-optimization","marketing",{"commitSha":264},{"parentExtensionId":538,"repoId":515},"k170sws65f0ebecn36z3q8c2z186m477",[222,534,455,224,535],{"evaluatedAt":541,"extractAt":519,"updatedAt":541},1778684020423,{"_creationTime":543,"_id":544,"community":545,"display":546,"identity":552,"providers":556,"relations":563,"tags":566,"workflow":567},1778695548458.3943,"k170qwsw3sjhsnm2r2nyz0j7jd86n1s7",{"reviewCount":8},{"description":547,"installMethods":548,"name":550,"sourceUrl":551},"Design and execute A/B tests for ML models in production using traffic splitting, statistical significance testing, and canary/shadow deployment strategies. Measure performance differences and make data-driven decisions about model rollout. Use when validating a new model version before full rollout, comparing candidate models trained with different algorithms, measuring business metric impact of model changes, or when regulatory requirements mandate gradual rollout.\n",{"claudeCode":549},"pjt222/agent-almanac","run-ab-test-models","https://github.com/pjt222/agent-almanac",{"basePath":553,"githubOwner":554,"githubRepo":555,"locale":18,"slug":550,"type":258},"skills/run-ab-test-models","pjt222","agent-almanac",{"evaluate":557,"extract":562},{"promptVersionExtension":214,"promptVersionScoring":215,"score":558,"tags":559,"targetMarket":227,"tier":228},95,[482,222,560,561,224],"canary-deployment","statistical-testing",{"commitSha":264},{"parentExtensionId":564,"repoId":565},"k170h0janaa9kwn7cfgfz2ykss86mmh9","kd7aryv63z61j39n2td1aeqkvh86mh12",[222,560,224,482,561],{"evaluatedAt":568,"extractAt":569,"updatedAt":568},1778700953339,1778695548458,{"_creationTime":571,"_id":572,"community":573,"display":574,"identity":580,"providers":584,"relations":592,"tags":594,"workflow":595},1778695660274.5056,"k17erc36epeh2pmf9zgrfhe19s86ne9f",{"reviewCount":8},{"description":575,"installMethods":576,"name":578,"sourceUrl":579},"Guides agents through the 3-step experiment creation flow: defining the hypothesis, configuring rollout, and setting up analytics. Delegates rollout decisions to configuring-experiment-rollout and metric setup to configuring-experiment-analytics.\nTRIGGER when: user asks to create a new experiment or A/B test, OR when you are about to call experiment-create.\nDO NOT TRIGGER when: user is updating an existing experiment, managing lifecycle, or only browsing experiments.",{"claudeCode":577},"PostHog/posthog","creating-experiments","https://github.com/PostHog/posthog",{"basePath":581,"githubOwner":582,"githubRepo":583,"locale":18,"slug":578,"type":258},"products/experiments/skills/creating-experiments","PostHog","posthog",{"evaluate":585,"extract":591},{"promptVersionExtension":214,"promptVersionScoring":215,"score":586,"tags":587,"targetMarket":227,"tier":590},79,[588,222,589,583,223],"experiments","product-analytics","community",{"commitSha":264},{"repoId":593},"kd7f22zf7qb3eschtk9s2qdv4586mfts",[222,588,223,583,589],{"evaluatedAt":596,"extractAt":597,"updatedAt":596},1778697075518,1778695660274]