# Beatnik Hiway > Literary Places & Counterculture Evidence API for source-backed cultural geography and clearly bounded visit planning. ## What this service does Beatnik Hiway exposes one published dataset snapshot through human-readable pages, JSON REST endpoints, Markdown, and a Streamable HTTP MCP endpoint. Its initial reviewed-production scope is San Francisco. Demonstration environments contain only clearly labelled synthetic records. ## Six tools - search_cultural_entities: Search the published cultural entity index (GET /api/v1/entities; cost 1 units) - get_entity_evidence: Retrieve claims and evidence for one entity (GET /api/v1/entities/{entity_id}; cost 1 units) - check_place_visitability: Evaluate scheduled access at a specified time (POST /api/v1/visitability/check; cost variable units) - build_cultural_walk: Build an evidence- and time-constrained cultural walk (POST /api/v1/walks/plan; cost 10 units) - build_research_pack: Build a deterministic, source-backed research pack (POST /api/v1/research-packs/build; cost 20 units) - get_dataset_coverage: Inspect current dataset scope and gaps (GET /api/v1/coverage; cost 1 units) ## Important endpoints - Documentation: https://beatnikhiway.com/docs - OpenAPI: https://beatnikhiway.com/openapi.json - Remote MCP: https://beatnikhiway.com/mcp - ARD status: https://beatnikhiway.com/.well-known/ard.json - A2A Agent Card status: https://beatnikhiway.com/.well-known/agent-card.json - Coverage: https://beatnikhiway.com/api/v1/coverage - Sources: https://beatnikhiway.com/data-sources - Methodology: https://beatnikhiway.com/methodology - Rights: https://beatnikhiway.com/rights - Health and current dataset: https://beatnikhiway.com/healthz ## Data sources The reviewed production snapshot uses official San Francisco DataSF records. Public source metadata and links: https://beatnikhiway.com/data-sources. ## Data freshness The current dataset version and publication state are reported by https://beatnikhiway.com/healthz. Each API response exposes dataset_version, freshness.reviewed_at, and served_at; the published snapshot is not a real-time venue or ticketing feed. ## Freshness and evidence Every response separates source publication, retrieval, review, validity, and serve times. Scheduled access is not live availability. Missing records mean not found in this dataset, not nonexistent in reality. ## Authentication and cost Small public reads are anonymous. Browser-rendered HTML uses a separate 2000-unit monthly page budget so machine API traffic cannot lock the website navigation; REST and MCP anonymous machine calls use the configured 500-unit monthly budget. Optional Bearer API keys provide higher quotas. Units are service quotas, not charges: search/evidence/coverage 1, visitability 1 per distinct place, walk 10, research pack 20. ## FAQ Common scope, freshness, source, authentication, action-limit, failure, and reuse questions: https://beatnikhiway.com/faq ## Limitations No ticketing, purchases, email, model calls, live maps, arbitrary URL fetching, OAuth, A2A Agent Card, ARD, or webhook task system is implemented in this release. The ARD and Agent Card paths are linked for discovery and return explicit machine-readable 404 responses. Training crawler access is disabled by configuration.