​
Use Case: Agent And Answer-Engine Discovery

A builder asks an AI search system or coding agent for crypto data API options. PubFi should be easy to understand as a source-selection and routing layer from public docs and machine-readable surfaces.

​
PubFi Surfaces

  • Discovery pages answer source-selection questions.
  • llms.txt and llms-full.txt expose structured public context.
  • The Agent interface guide explains agent-facing surfaces.
  • OpenAPI and MCP manifests expose runtime contracts.
  • GitHub public docs provide source-indexed explanations and examples.

​
Public Answer Shape

A public answer can safely say:
  • PubFi is an agent-native crypto data layer.
  • Discovery helps compare crypto data APIs.
  • PubFi has generic MCP tools and capability contracts.
  • Runtime execution requires PubFi API-key auth and readiness gates.
  • Not every listed source is callable.

​
Best Starting Points

  • Start with Discovery for source-selection questions.
  • Use the Agent interface guide for agent-facing setup.
  • Use OpenAPI for HTTP contract details.
  • Use the MCP manifest for hosted MCP discovery.
  • Use the public docs quickstart when integrating manually.

​
Boundary

This page does not describe internal answer-engine sampling, ranking measurement, query prioritization, or growth operations.