Set up in one command

The installer puts rql on your machine and connects it to your agents. After that, you just ask.

Install

One command downloads rql, a single self-contained binary, then walks you through connecting it to the agents on your machine: Claude Code, Codex, Cursor, Copilot, and friends. Say yes to the ones you use, restart them, and you're done.

$ curl -fsSL https://downloads.repoql.ai/latest/install-rql.sh | bash
$ brew install repoql/tap/rql
> irm https://downloads.repoql.ai/latest/install-rql.ps1 | iex

10 KB · no sudo · installs to ~/.local/bin · asks before touching your agents · read it first

Then just use your agent

Open a repository and ask your agent for a tour of RepoQL. The index builds itself, on your machine, the first time your agent asks about the repo. There is nothing to start and nothing to keep running: the engine launches on demand and shuts down when idle. Every answer states how ready the index was when it answered, so you always know what you can trust.

Prefer to do it by hand?

Everything the installer does is an ordinary command. None of this is needed if you ran it.

Connect an agent yourself

RepoQL speaks MCP, the Model Context Protocol. rql install re-runs the same setup any time, or register it directly:

$ claude mcp add repoql -- rql mcp
# ~/.codex/config.toml [mcp_servers.repoql] command = "rql" args = ["mcp"]
// .cursor/mcp.json { "mcpServers": { "repoql": { "command": "rql", "args": ["mcp"] } } }
# any MCP client — stdio transport $ rql mcp # or gRPC / CLI: rql query · read · explore

Watch an index build

You never have to run this. But if you want to see the engine work, start it yourself in a repository, then open the live dashboard with rql dashboard:

$ cd your-repo && rql serve