Summary
The API suite your team already maintains becomes something an agent can run and read.
Collection, environment and variables are all arguments, so nothing about the suite has to change to make it callable, and the same collection can be pointed at a different target by name. Authoring stays in Bruno: there is no tool here for editing a request, inspecting one response in isolation, or managing environments, and the result you get is whatever the CLI run produced.
What it is
A single-tool wrapper over the Bruno CLI that executes a collection you point it at and returns what the run produced.
What you get
- A Bruno collection executed by path, so the request suite already sitting in the repository is what runs
- Environment selection per run through the environment argument, without editing the collection to switch targets
- Per-run overrides through the variables argument, for a base URL or a token that differs between one run and the next
Requirements
A Bruno collection on the machine and the Bruno CLI available to execute it. No account, no key and no environment variables of its own.
Setup effort
One command — npx -y bruno-mcp