About container-use
Container Use is an open-source MCP server that gives coding agents isolated containerized development environments. It works as a CLI tool alongside Claude Code, Cursor and other MCP-compatible agents. Each agent works on its own git branch, which suits running several agents in parallel. The project is built on Dagger and is in early development.
Key features
- 01Isolated container environment and git branch per agent
- 02Complete command history and logs of agent actions
- 03Direct access to any agent's terminal for intervention
- 04Review agent work through standard git checkout
- 05Compatible with any agent, model or infrastructure
Use cases
- 01Running several coding agents on one project without conflicts
- 02Trying changes in isolation and discarding failed attempts
- 03Checking the commands and logs an agent actually executed
MCP support
The project runs as an MCP server, and MCP clients add it by running container-use stdio.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.
Tools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project README
container-use
Containerized environments for coding agents. (📦🤖) (📦🤖) (📦🤖)
Container Use lets coding agents do their work in parallel environments without getting in your way. Go from babysitting one agent at a time to enabling multiple agents to work safely and independently with your preferred stack. See the full documentation.
It's an open-source MCP server that works as a CLI tool with Claude Code, Cursor, and other MCP-compatible agents. Powered by Dagger.
- 📦 Isolated Environments: Each agent gets a fresh container in its own git branch - run multiple agents without conflicts, experiment safely, discard failures instantly.
- 👀 Real-time Visibility: See complete command history and logs of what agents actually did, not just what they claim.
- 🚁 Direct Intervention: Drop into any agent's terminal to see their state and take control when they get stuck.
- 🎮 Environment Control: Standard git workflow - just
git checkout <branch_name>to review any agent's work. - 🌎 Universal Compatibility: Works with any agent, model, or infrastructure - no vendor lock-in.
🦺 This project is in early development and actively evolving. Submit issues and/or reach out to us on Discord in the #container-use channel.
Quick Start
Install
# macOS (recommended)
brew install dagger/tap/container-use
# All platforms
curl -fsSL https://raw.githubusercontent.com/dagger/container-use/main/install.sh | bash
Setup with Your Agent
Container Use works with any MCP-compatible agent. The setup is always the same: add container-use stdio as an MCP server.
👉 Complete setup guide for all agents (Cursor, Goose, VSCode, etc.)
Example with Claude Code:
# Add Container Use MCP server
cd /path/to/repository
claude mcp add container-use -- container-use stdio
# Add agent rules (optional)
curl https://raw.githubusercontent.com/dagger/container-use/main/rules/agent.md >> CLAUDE.md
💡 Command Shortcut
The container-use command is also available as cu for convenience. Both commands work identically:
container-use stdio(used in documentation)cu stdio(shortcut)
Try It
Ask your agent to create something:
Create a hello world app in python using flask
Your agent will work in an isolated environment and give you URLs to view the app and explore the code!
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.