This documentation is built to be read by AI assistants and coding agents. Feed it the whole site as text, connect it as a documentation MCP server, or grab any page as Markdown.

llms.txt#

Following the llms.txt convention, two machine-readable files expose the docs as plain text:

• llms.txt: a compact, link-first index of every page and endpoint.
• llms-full.txt: the entire documentation as one text payload, ready to paste into a model's context.

MCP documentation server#

The docs are also an MCP server at https://routescan.io/docs/mcp. Connect it in Cursor, Claude, VS Code or any MCP client and the assistant can search and read these docs directly while it works. It exposes three tools:

• search_routescan: ranked search across the documentation, returning page titles, URLs and snippets.
• find_endpoints: the endpoints that match a goal, as structured JSON with their parameters and a ready-to-run example.
• query_docs_filesystem_routescan: a read-only shell (ls, tree, cat, grep/rg) over the docs as Markdown files, including the full API reference.

This is the documentation MCP: it searches and reads the docs. It is a different server from Routescan's on-chain data MCP server (mcp.routescan.io/mcp, see Build with AI), which answers questions about blockchain data, not the docs.

The quickest way to install is the Copy page menu at the top of any page (Connect to Cursor / Connect to VS Code). To add it manually, use:

Per-page Markdown#

Append .md to any API reference URL for its raw Markdown (for example /docs/api/addresses.md), or use the Copy page button to copy the current page. Selecting everything on a page and copying also yields Markdown rather than rendered text.