pubfi.capabilities.search | Search PubFi’s capability catalog for routeable crypto data needs. This returns read-only catalog metadata and does not call upstream providers or spend credits. | query, required_capabilities, categories |
pubfi.route.plan | Create a dry-run PubFi route plan from an intent, objective, chains, categories, or required capabilities. This is read-only and does not call upstream providers, execute routes, or spend credits. | intent, objective, chains, categories, required_capabilities, dry_run |
pubfi.route.execute | Execute a selected PubFi capability route using route-planning evidence from pubfi.route.plan. This requires a PubFi API key, may consume account credits, and rejects provider-specific route ids outside the generic capability surface. | route_id, route_plan, arguments, idempotency_key, request_id |
pubfi.route.explain | Explain the routing decision PubFi would make for an intent. This read-only diagnostic returns reason codes, filters, and selected route metadata without executing providers or spending credits. | intent, objective, chains, categories, required_capabilities, dry_run |
pubfi.schema.get | Return PubFi MCP input and output schema details for a named tool, especially pubfi.route.execute. This is read-only and intended for agent setup and validation before planning or execution. | tool |
pubfi.pricing.quote | Estimate PubFi credits and readiness for a capability route. This read-only quote does not create payment payloads, procure supplier access, or execute the route. | route_id, capability_id, intent |