transport
Streamable HTTP MCP
A stateless Model Context Protocol server at https://api.refd.ai/mcp.
agent access
Build with refd. Read your AI visibility from any agent.
refd exposes a workspace's AI search monitoring through a read-only, OAuth-protected remote MCP server. Connect Claude, ChatGPT, or any MCP client and query visibility, competitors, citations, and the raw answers behind them, scoped to a single workspace.
01 / the connector
A read-only connector, scoped to one workspace.
Every OAuth grant is bound to a single workspace and is read-only. The connector cannot change data or start paid runs, and the owner can revoke it anytime from Settings. Web prompt-injection can, at worst, read within the one workspace the human authorized.
auth
OAuth 2.1 + PKCE
Authorization-code flow with per-workspace, revocable grants and the data:read scope.
access
Read-only
Nine tools plus a metric-glossary resource. No writes, no run triggers, no spend.
02 / the tools
Nine read tools, discovered at runtime.
Call tools/list after connecting. Every tool resolves the workspace from the OAuth grant, never from arguments.
get_workspace_info
The brand, tracked competitors, prompts, and enabled AI surfaces for the workspace.
get_visibility_overview
Mention rate, citation rate, position, and share of voice across surfaces.
get_competitor_landscape
How the brand ranks against the competitors it tracks.
get_prompt_performance
Per-prompt visibility, broken down by AI surface.
get_citation_sources
Which domains AI answers cite for the workspace.
get_recent_changes
Material moves between the two most recent completed runs.
find_prompt_results
Search tracked prompts and their scored results.
read_answer
The raw AI answer behind a result, with entity mentions highlighted.
get_digest
A 30-day rollup of the workspace, the same one that grounds the dashboard chat.
03 / connect
Point a client at the endpoint.
The endpoint is always https://api.refd.ai/mcp. Discovery and sign-in are automatic; you authorize one workspace per connection.
Claude / Claude Code
Add a custom connector (or `claude mcp add --transport http refd https://api.refd.ai/mcp`) and complete the OAuth sign-in.
ChatGPT
Settings → Connectors → add a custom MCP server, enter the endpoint, and authorize.
Any MCP client
Point a Streamable HTTP MCP client at the endpoint; it discovers auth via the protected-resource metadata.
04 / discovery
Everything is discoverable.
MCP endpoint
https://api.refd.ai/mcp
Streamable HTTP, OAuth-protected
Protected-resource metadata
https://api.refd.ai/.well-known/oauth-protected-resource/mcp
RFC 9728
Authorization-server metadata
https://api.refd.ai/.well-known/oauth-authorization-server
RFC 8414
OpenAPI catalog
https://refd.ai/openapi.json
Public HTTP surface
Agent manifest
https://refd.ai/.well-known/agent
Discovery pointers
MCP Registry
ai.refd/refd
registry.modelcontextprotocol.io
llms.txt
https://refd.ai/llms.txt
Plain-text summary
05 / open source
Self-host the whole stack.
refd is MIT licensed and runs on Cloudflare Workers. Run the MCP server on your own account, or connect to the hosted one on refd.ai.