Docker Agent is Docker's open-source framework for putting together teams of AI agents that are described in a YAML file and run from the terminal.
In plain words
Instead of asking one general model to do everything, Docker Agent lets you split a job across several agents, each with its own role, model, instructions and tools, such as file access, a shell or MCP servers. A root agent can hand parts of the job to sub-agents. You write the team in a YAML file and start it with the docker agent command, using models from many providers. Docker Desktop 4.63 and later include it; earlier versions called it cagent. The code is under the Apache 2.0 licence.
A quick example
Imagine a two-agent bug team. The first agent reads an error message and the code, works out the cause, then passes the job to a second agent, which makes a small fix and adds a test so the bug cannot quietly return.
Where you'll meet it
- Coding helpers that split debugging, fixing and testing between agents.
- Agent teams started from a terminal or included with Docker Desktop.
Related
Sources
- DocsDocker Agent, Docker Docs · read 7 Oct 2026
- Docsdocker/docker-agent, Docker (GitHub) · read 7 Oct 2026
Checked by WhatIsAI against the sources above · 7 Oct 2026. A full five-minute explainer with a diagram may follow. Spotted an error? Report it.