Documentation
Build with the readiness network
Everything is an HTTP endpoint or an MCP tool. Three calls — search, preflight, outcome — close the loop.
Quickstart
Search for a capability, preflight it, and report an outcome in under five minutes.
Search API
Turn a task description into ranked, evidence-backed capability candidates.
Preflight API
A last-second readiness check before your agent commits to a call.
Outcome Feedback API
Report whether the result was actually usable, and improve scoring for every agent.
MCP Server
Use the network as an MCP server from Claude Code, Cursor, Cline, and other MCP clients.
Capability Manifest
The submission format for providers: interfaces, actions, and risk annotations.
Evidence Levels
E0–E7: how much verification stands behind every score.