What To Read First
- Project overview: what PubFi is and what it is not.
- Quickstart: the shortest path through public docs and live surfaces.
- API reference: where to find the interactive HTTP reference and OpenAPI schema.
- API key and runtime: how account, credit, and runtime boundaries fit together.
- MCP client setup: connect an agent runtime to PubFi’s generic MCP tools.
- Discovery: understand the public source-selection layer.
- Capability contracts: understand normalized agent-facing responses.
- Agent-readable surfaces: understand
llms.txt,llms-full.txt, OpenAPI, MCP, and Discovery links.
Public Surfaces
| Surface | Purpose |
|---|---|
| Product | product, account, Discovery, and agent-readable product entry |
| Discovery | public crypto data API index and source-selection surface |
| llms.txt | concise site index for LLM and crawler discovery |
| llms-full.txt | expanded Discovery export for answer-engine retrieval |
| API Reference | interactive API reference |
| OpenAPI | API schema |
| MCP manifest | MCP discovery manifest |
Claim Boundary
Discovery inclusion does not mean gateway availability. A route plan does not authorize execution. Local SEO/GEO evidence does not prove ranking, traffic, or AI citation success. PubFi keeps those boundaries explicit so agents can choose routes safely.Coverage Map
| Area | Start here |
|---|---|
| Product positioning | Project overview |
| HTTP API details | API reference |
| API key, account, and credits | API key and runtime |
| Agent/MCP interface | MCP client setup |
| Discovery and source selection | Discovery |
| Provider readiness and certification | Provider readiness |
| Agent-readable surfaces | Agent-readable surfaces |
| Runtime endpoints | Runtime endpoints |
| Public/security boundary | Security and public data |