{"passport":{"unfragile":{"@version":"1.0","version":"2026-05","artifact":{"id":"npm_npm-growthbookmcp","slug":"npm-growthbookmcp","name":"@growthbook/mcp","type":"mcp","url":"https://www.npmjs.com/package/@growthbook/mcp","page_url":"https://unfragile.ai/npm-growthbookmcp","categories":["mcp-servers"],"tags":["growthbook","mcp","modelcontextprotocol","featureflags","experiments"],"pricing":{"model":"open_source","free":true,"starting_price":null},"status":"active","verified":false},"capabilities":[{"id":"npm_npm-growthbookmcp__cap_0","uri":"capability://tool.use.integration.growthbook.feature.flag.evaluation.via.mcp.protocol","name":"growthbook feature flag evaluation via mcp protocol","description":"Exposes GrowthBook's feature flag evaluation engine through the Model Context Protocol, allowing Claude and other MCP-compatible clients to query feature flag states, variations, and targeting rules without direct API calls. Implements MCP resource and tool handlers that translate client requests into GrowthBook SDK method calls, maintaining session context across multiple evaluations within a single conversation.","intents":["Query whether a specific feature flag is enabled for a given user context","Retrieve the assigned variation for an experiment or feature flag","Evaluate feature flags with custom user attributes and context","Integrate feature flag checks into AI-driven decision workflows"],"best_for":["AI agents and Claude instances that need to make feature-flag-aware decisions","Teams building AI-powered feature management dashboards","Developers integrating GrowthBook into LLM-based workflows"],"limitations":["Requires GrowthBook SDK initialization before MCP server startup — no lazy loading of credentials","Evaluation results are point-in-time snapshots; real-time flag changes require re-evaluation","No built-in caching of evaluation results across MCP calls, each query hits GrowthBook SDK"],"requires":["GrowthBook account with API credentials or SDK key","Node.js 16+ runtime","@growthbook/sdk package installed","MCP-compatible client (Claude, or custom MCP host)"],"input_types":["feature flag identifier (string)","user context object (JSON with userId, attributes, etc.)","experiment/variation identifiers"],"output_types":["boolean (flag enabled/disabled)","string or JSON (variation value)","structured evaluation result with rule metadata"],"categories":["tool-use-integration","feature-management"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm_npm-growthbookmcp__cap_1","uri":"capability://tool.use.integration.experiment.metadata.and.targeting.rule.retrieval","name":"experiment metadata and targeting rule retrieval","description":"Provides MCP tools to fetch experiment configurations, targeting rules, and variation assignments from GrowthBook without requiring direct REST API calls. Implements resource handlers that serialize GrowthBook experiment objects into structured JSON, exposing rule conditions, audience targeting, and variation weights for inspection and decision-making within AI workflows.","intents":["Retrieve full experiment configuration including targeting rules and variation details","Inspect audience targeting conditions to understand who qualifies for an experiment","List all active experiments and their current status","Examine variation weights and assignment logic for an experiment"],"best_for":["AI agents that need to reason about experiment eligibility before making decisions","Teams building experiment analysis or audit tools powered by Claude","Developers debugging feature flag and experiment targeting logic"],"limitations":["Metadata is read-only; no capability to create or modify experiments through MCP","Targeting rule serialization may be lossy for complex custom rules not natively supported by GrowthBook SDK","No pagination or filtering — large experiment lists are returned in full"],"requires":["GrowthBook SDK initialized with valid credentials","Read access to experiment metadata in GrowthBook account","Node.js 16+"],"input_types":["experiment ID (string)","optional filter parameters (JSON)"],"output_types":["experiment configuration object (JSON)","targeting rules array (JSON)","variation metadata (JSON)"],"categories":["tool-use-integration","data-processing-analysis"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm_npm-growthbookmcp__cap_2","uri":"capability://tool.use.integration.user.context.and.attribute.based.flag.evaluation","name":"user context and attribute-based flag evaluation","description":"Enables MCP clients to evaluate feature flags with arbitrary user attributes and context, implementing a schema-based parameter handler that maps user context objects to GrowthBook SDK evaluation calls. Supports custom attributes, user IDs, and environment-specific context, allowing Claude to simulate flag behavior for different user segments without hardcoding evaluation logic.","intents":["Evaluate a feature flag for a hypothetical user with specific attributes","Test how a flag behaves across different user segments or personas","Simulate feature flag behavior for debugging or decision-making","Build AI workflows that adapt behavior based on user context and flag state"],"best_for":["AI agents that need to reason about feature flag behavior for different user cohorts","Teams building personalization or targeting logic with AI assistance","Developers testing flag behavior without manual SDK integration"],"limitations":["Attribute validation is delegated to GrowthBook SDK; invalid attributes may be silently ignored","No support for real-time attribute updates — context is static per evaluation call","Custom attribute types must be JSON-serializable; complex objects may not round-trip correctly"],"requires":["GrowthBook SDK initialized","Knowledge of valid attribute names and types for your GrowthBook account","Node.js 16+"],"input_types":["user context object: { userId, attributes: { key: value, ... }, environment?: string }","feature flag identifier"],"output_types":["boolean (flag state)","variation value (string, number, JSON)","evaluation metadata (rule matched, timestamp)"],"categories":["tool-use-integration","data-processing-analysis"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm_npm-growthbookmcp__cap_3","uri":"capability://tool.use.integration.mcp.resource.exposure.for.feature.flags.and.experiments","name":"mcp resource exposure for feature flags and experiments","description":"Registers GrowthBook feature flags and experiments as MCP resources, making them discoverable and queryable by Claude through the MCP resource protocol. Implements resource URI schemes (e.g., growthbook://flag/{id}, growthbook://experiment/{id}) that map to GrowthBook SDK objects, allowing Claude to reference and inspect flags/experiments as first-class entities within conversations.","intents":["Discover available feature flags and experiments in GrowthBook","Reference a specific flag or experiment by URI in a conversation","Inspect flag/experiment metadata as a resource without explicit tool calls","Build Claude prompts that reference GrowthBook resources by URI"],"best_for":["Teams using Claude with MCP where flags/experiments should be first-class conversation entities","Developers building Claude-based feature management interfaces","AI workflows that need to reference GrowthBook resources by stable identifiers"],"limitations":["Resource discovery is static at MCP server startup; new flags/experiments require server restart","Resource URIs are read-only; no mutation through resource protocol","Large numbers of flags/experiments may impact MCP resource enumeration performance"],"requires":["GrowthBook SDK initialized","MCP-compatible client that supports resource protocol","Node.js 16+"],"input_types":["resource URI (growthbook://flag/{id} or growthbook://experiment/{id})"],"output_types":["resource metadata (JSON)","flag/experiment configuration object"],"categories":["tool-use-integration","memory-knowledge"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm_npm-growthbookmcp__cap_4","uri":"capability://tool.use.integration.sdk.initialization.and.credential.management.via.mcp","name":"sdk initialization and credential management via mcp","description":"Handles GrowthBook SDK initialization and credential management at MCP server startup, accepting configuration through environment variables or constructor parameters. Implements a single-instance SDK pattern where credentials are loaded once and reused across all MCP tool/resource calls, eliminating per-request authentication overhead and maintaining consistent evaluation state.","intents":["Initialize GrowthBook SDK with API key or SDK key at server startup","Manage credentials securely without exposing them in MCP messages","Ensure all MCP calls use the same authenticated SDK instance","Configure SDK options (caching, refresh intervals) once for all clients"],"best_for":["Teams deploying MCP servers in secure environments where credentials can be injected via env vars","Developers building self-hosted MCP servers for GrowthBook","Organizations that need centralized credential management for multiple Claude instances"],"limitations":["Credentials must be provided at server startup; no runtime credential rotation","Single SDK instance means all MCP clients share the same evaluation cache and state","No support for multi-tenant scenarios where different clients need different GrowthBook accounts"],"requires":["GrowthBook API key or SDK key (environment variable: GROWTHBOOK_API_KEY or GROWTHBOOK_SDK_KEY)","Node.js 16+","GrowthBook account with valid credentials"],"input_types":["environment variables or constructor config object"],"output_types":["initialized GrowthBook SDK instance (internal)"],"categories":["tool-use-integration","safety-moderation"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm_npm-growthbookmcp__cap_5","uri":"capability://tool.use.integration.variation.assignment.and.a.b.test.result.querying","name":"variation assignment and a/b test result querying","description":"Provides MCP tools to query which variation a user is assigned to in an A/B test and retrieve test results/metrics from GrowthBook. Implements evaluation logic that returns both the assigned variation and associated metadata (rule matched, timestamp, experiment ID), enabling Claude to understand test outcomes and make data-driven decisions based on experiment results.","intents":["Determine which variation a specific user is assigned to in an experiment","Retrieve A/B test results and statistical significance data","Query experiment metrics and conversion rates","Build AI workflows that adapt based on test outcomes"],"best_for":["Teams building AI-powered experiment analysis and reporting tools","Developers integrating A/B test results into Claude-based decision workflows","Product managers using Claude to interpret experiment outcomes"],"limitations":["Variation assignment is deterministic based on user ID and experiment rules; no support for random assignment","Metrics/results querying depends on GrowthBook's analytics integration — may require additional setup","No support for multi-armed bandit or adaptive allocation algorithms"],"requires":["GrowthBook SDK initialized","Active experiments with defined variations","Node.js 16+"],"input_types":["user ID (string)","experiment ID (string)","user attributes (optional, JSON)"],"output_types":["variation identifier (string)","variation value (JSON)","experiment metadata (rule matched, timestamp)","test results/metrics (if available)"],"categories":["tool-use-integration","data-processing-analysis"],"confidence":0.5,"matches":0,"success_rate":0}],"trust":{"score":30,"verified":false,"data_access_risk":"high","permissions":["GrowthBook account with API credentials or SDK key","Node.js 16+ runtime","@growthbook/sdk package installed","MCP-compatible client (Claude, or custom MCP host)","GrowthBook SDK initialized with valid credentials","Read access to experiment metadata in GrowthBook account","Node.js 16+","GrowthBook SDK initialized","Knowledge of valid attribute names and types for your GrowthBook account","MCP-compatible client that supports resource protocol"],"failure_modes":["Requires GrowthBook SDK initialization before MCP server startup — no lazy loading of credentials","Evaluation results are point-in-time snapshots; real-time flag changes require re-evaluation","No built-in caching of evaluation results across MCP calls, each query hits GrowthBook SDK","Metadata is read-only; no capability to create or modify experiments through MCP","Targeting rule serialization may be lossy for complex custom rules not natively supported by GrowthBook SDK","No pagination or filtering — large experiment lists are returned in full","Attribute validation is delegated to GrowthBook SDK; invalid attributes may be silently ignored","No support for real-time attribute updates — context is static per evaluation call","Custom attribute types must be JSON-serializable; complex objects may not round-trip correctly","Resource discovery is static at MCP server startup; new flags/experiments require server restart","builder identity is not verified yet","no observed match outcomes yet"],"rank_breakdown":{"adoption":0.05,"quality":0.22,"ecosystem":0.45,"match_graph":0.25,"freshness":0.9,"weights":{"adoption":0.25,"quality":0.25,"ecosystem":0.15,"match_graph":0.23,"freshness":0.12}},"observed_outcomes":{"matches":0,"success_rate":0,"avg_confidence":0,"top_intents":[],"last_matched_at":null},"maintenance":{"status":"active","updated_at":"2026-05-24T12:16:23.903Z","last_scraped_at":"2026-05-03T14:23:54.079Z","last_commit":null},"community":{"stars":null,"forks":null,"weekly_downloads":null,"model_downloads":null,"model_likes":null}},"distribution":{"claim_url":"https://unfragile.ai/submit?claim=npm-growthbookmcp","compare_url":"https://unfragile.ai/compare?artifact=npm-growthbookmcp"}},"signature":"Jd0u5tKmPM38bS8WYLNi3HZEUbnrgkAQfe8IdPTdXeQeQFbz/dWkWDftVlqfoHcm7/+lv2DqU4xx7ax0ROGoCQ==","signedAt":"2026-06-17T01:33:59.785Z","signedBy":"unfragile.ai","version":1},"_links":{"self":"https://unfragile.ai/api/v1/passport/npm-growthbookmcp","artifact":"https://unfragile.ai/npm-growthbookmcp","verify":"https://unfragile.ai/api/v1/verify?slug=npm-growthbookmcp","publicKey":"https://unfragile.ai/api/v1/trust-passport-public-key","spec":"https://unfragile.ai/trust","schema":"https://unfragile.ai/schema.json","docs":"https://unfragile.ai/docs"}}