Reference
What is MCP?
The Model Context Protocol is an open standard that lets AI clients use your tools and data. SlashFeed speaks it — so your reading list becomes something your own AI can actually work with.
MCP (Model Context Protocol) is an open protocol for connecting AI assistants to external tools and data sources. Instead of copy-pasting into a chat, an MCP-capable client (like Claude, Cursor, or Zed) connects to a server that exposes tools it can call and resources it can read — securely and on your behalf.
How it works
- A server exposes tools (e.g. "search my entries") over MCP.
- A client (your AI app) connects and authenticates.
- The model can then call those tools — reading, searching, and acting — grounded in your real data rather than guessing.
SlashFeed's MCP server
Every account gets a personal MCP endpoint, scoped to your own feeds and secured with a personal token or OAuth 2.0 + PKCE:
MCP endpoint: https://slashfeed.app/api/mcp Auth: personal token, or OAuth 2.0 + PKCE
Point an MCP client at it and the model can, on your behalf:
- list and search your entries and feeds;
- mark entries read and save them for later;
- read your reading statistics.
Unlike a built-in, black-box assistant, you bring your own model — nothing is locked to one AI vendor. See the API & MCP documentation for the full tool list.
Frequently asked questions
Which AI clients support MCP?
A growing set — including Claude (desktop and web), Cursor, and Zed. Any client that implements the Model Context Protocol can connect.
Is my data safe with MCP?
Your endpoint is scoped to your own account and authenticated. You grant access explicitly and can revoke the token at any time. See the security page.
Do I need to pay for AI features?
No SlashFeed "AI tier" — MCP access is included, and you use whichever model your own client provides.