{"id":"lzinga-us-gov-open-data-mcp","name":"us-gov-open-data-mcp","homepage":"https://lzinga.github.io/us-gov-open-data-mcp/","repo_url":"https://github.com/lzinga/us-gov-open-data-mcp","category":"api-gateway","subcategories":[],"tags":["mcp","model-context-protocol","open-data","government","typescript","sdk","data-integration","caching","rate-limiting"],"what_it_does":"Provides an MCP server (via stdio and HTTP Stream) and a TypeScript SDK that expose 300+ tools across 40+ U.S. government (and some international) open-data APIs, with module-level selective loading, disk-backed caching, and built-in rate limiting/retry.","use_cases":["Answering questions about U.S. government data (economy, health, legislation, spending, environment, etc.) via an AI assistant","Cross-referencing multiple agencies’ datasets (e.g., FDA + lobbying + campaign finance)","Programmatic data retrieval in TypeScript without using MCP","Interactive exploration of government datasets through MCP-enabled desktop clients"],"not_for":["Use as a data warehouse or to persistently host authoritative datasets","Scenarios requiring custom compliance/validation workflows beyond what upstream government APIs provide","Environments that cannot allow local execution (npx) or require fully managed hosted endpoints"],"best_when":"You want broad coverage of U.S. government APIs accessible to AI agents or typed code, with caching and rate-limit handling, and you can supply any needed upstream API keys via environment variables.","avoid_when":"You need a standardized single REST/OpenAPI contract for all tools, strict enterprise security controls (e.g., no local process execution), or strong guarantees about per-endpoint schema correctness beyond the stated community/accuracy caveats.","alternatives":["Direct use of each underlying government API (e.g., FRED, USAspending, Congress.gov APIs) with your own client code","General-purpose MCP data connectors from other vendors","Fewer-purpose API aggregators focused on one agency/domain"],"af_score":77.0,"security_score":60.0,"reliability_score":31.2,"package_type":"mcp_server","discovery_source":["github"],"priority":"high","status":"evaluated","version_evaluated":null,"last_evaluated":"2026-03-30T13:48:17.002887+00:00","interface":{"has_rest_api":false,"has_graphql":false,"has_grpc":false,"has_mcp_server":true,"mcp_server_url":null,"has_sdk":true,"sdk_languages":["TypeScript","JavaScript"],"openapi_spec_url":null,"webhooks":false},"auth":{"methods":["Environment variable API keys for upstream APIs (e.g., FRED_API_KEY, DATA_GOV_API_KEY)"],"oauth":false,"scopes":false,"notes":"Authentication is handled by passing upstream API keys via environment variables in the MCP client config. The README states 20+ APIs require no key; the rest use free keys (time-to-get under a minute), but it does not specify per-tool scope granularity."},"pricing":{"model":null,"free_tier_exists":false,"free_tier_limits":null,"paid_tiers":[],"requires_credit_card":false,"estimated_workload_costs":null,"notes":"The package appears MIT-licensed (code distribution). Pricing for upstream APIs (where applicable) is not described here."},"requirements":{"requires_signup":false,"requires_credit_card":false,"domain_verification":false,"data_residency":[],"compliance":[],"min_contract":null},"agent_readiness":{"af_score":77.0,"security_score":60.0,"reliability_score":31.2,"mcp_server_quality":78.0,"documentation_accuracy":72.0,"error_message_quality":0.0,"error_message_notes":null,"auth_complexity":75.0,"rate_limit_clarity":70.0,"tls_enforcement":85.0,"auth_strength":55.0,"scope_granularity":35.0,"dependency_hygiene":55.0,"secret_handling":70.0,"security_notes":"Uses environment variables for upstream API keys (better than embedding in code, and commonly not logged), and implements rate limiting/retry to reduce abusive request patterns. However, it relies on many third-party endpoints/schemas; the security strength of the aggregator itself (e.g., structured error handling, secret redaction guarantees, and TLS enforcement details) is not fully evidenced in the provided README/manifest content.","uptime_documented":0.0,"version_stability":55.0,"breaking_changes_history":0.0,"error_recovery":70.0,"idempotency_support":"false","idempotency_notes":null,"pagination_style":"unknown","retry_guidance_documented":true,"known_agent_gotchas":["Large number of tools (300+) increases the chance an agent calls an endpoint with mismatched parameters/schema; the README warns about incomplete parameter coverage/unexpected endpoint behaviors.","Some APIs require keys; missing/invalid env vars may cause tool failures.","Because selective loading exists, agents may need to ensure the relevant modules are enabled for the tools they intend to call.","Local npx execution and disk-backed caching may create environment-specific behavior (permissions/path) if running under constrained desktops/containers."]}}