Login Book a demo

MCP

The Ollive MCP server.
Your agents’ record, in conversation.

Connect Ollive to Claude, Cursor, Codex or Gemini CLI, and ask about your agents in your own words: which rules they broke, why, how often, and what was said. Your assistant finds the answer in your Ollive workspace and brings the evidence with it.

Server URL
https://mcp.ollive.ai/mcp
Sign-in method
OAuth, with your Ollive login
Transport
Streamable HTTP
Connect in 2 simple steps

Ask in your own words

There’s no query language to learn and no dashboard to dig through. Ask the way you’d ask a colleague, then follow up.

  • AskHow did my agents do this week?
  • AskWhich rules does the appointment booking agent break most often?
  • AskShow me the conversations behind that violation.
  • AskWhy did conversation call-2026-10-14-0912 fail?
  • AskAre high-severity violations going down this month?
  • AskSend these test conversations to Ollive and tell me what fails.

How it works

  1. Add Ollive to your app

    Give Claude, Cursor, Codex or Gemini CLI the server URL. There’s no API key to create or paste.

  2. Sign in and allow

    Your browser opens Ollive, and you sign in with OAuth. Choose the workspace, and select Allow.

Then just ask. Your assistant picks the Ollive tools it needs, reads your workspace and answers, with the evidence.

Choose your app

Claude

Add Ollive as a connector in Claude, on the web and in the desktop app.

Setup guide

Claude Code

Add Ollive to Claude Code with one command.

Setup guide

Cursor

Add Ollive to Cursor and use it in Agent chat.

Setup guide

Codex

Add Ollive to Codex, in the CLI, the IDE extension and the app.

Setup guide

Gemini CLI

Add Ollive to Gemini CLI from your terminal.

Setup guide

Other apps that support remote MCP servers with sign-in connect the same way: give them the server URL above.

What your assistant can do

Ollive gives your assistant seven tools. Six of them only read. The seventh sends conversations to be checked, and only when you ask it to. See all the tools.

Prefer to work in code?

The tools return the same results as the Ollive API. Use the API to build Ollive into your own systems, and MCP to ask questions as they come up.

← Docs homeQuickstart →