Slovey
MCP server

One memory for every agent.

Slovey speaks the Model Context Protocol, so any MCP-capable agent can read your decisions, rules and history before it writes a line.

Claude Code

Checks your decision graph before it acts on a change.

Cursor

Uses the same context your reviewers and CI use.

MCP

Codex and others

Anything that speaks MCP connects to the same memory.

Set it up

There is nothing to install or run. Add this address to your agent and it signs in with Slovey: you pick the repository it may read and click Allow. Access is scoped to that one repository, and you can remove it any time in Settings.

https://slovey.ai/api/mcp

In Claude (the app or claude.ai), open Settings, then Connectors, then Add custom connector, and paste it. In Claude Code it is one command, then /mcp to sign in:

claude mcp add --transport http slovey https://slovey.ai/api/mcp

Cursor and VS Code get a one-click install on the Connect page in your dashboard, which also shows the moment your agent first reaches Slovey.

Your agent then gets six tools: what_applies_here, plan, can_i, ask, get_rejected and check_change, which checks a diff before it is committed. Prefer a local server? npm i -g slovey ships slovey-mcp, which also runs Preflight against your working tree.

Prefer HTTP? The same answers are on the API. Running Slovey yourself? Set DATABASE_URL and SLOVEY_REPO instead of a token; see the MCP readme.

Give your agents your team's memory.