# MCO

A lightweight, CLI-first orchestration layer that runs a chosen set of AI coding agents and models on the same task in parallel, keeps each answer opaque and separate, and lets a developer compare them before acting.

MCO dispatches one task to the agents and models a developer selects, runs them in parallel, and retains each invocation's complete raw answer and operational status rather than merging them into a single automatic decision. It is meant for code review, implementation, architecture analysis and CI checks where one model's blind spots matter.

It never infers a provider or model team from the binaries it detects on a machine; the team must be selected explicitly, through a `--providers` flag, a saved configuration default, or a calling agent that confirms the resolved team with the user before dispatch. A `--perspectives-json` option can give each provider a different prompt focus, and `--divide files` or `--divide dimensions` splits a review across files or across review lenses.

It runs from a terminal or from another coding agent such as Claude Code, Codex, Cursor, Copilot, Pi or OpenClaw, and installs its CLI and a bundled Skill with `npx @tt-a1i/mco@latest install`.

Website: https://github.com/mco-org/mco