AI coding agents are useful the moment they can reach your infrastructure - link a project, deploy it, run a pipeline, read the build logs. Buddy connects agents to all of that in two directions: an agent running on your machine can drive Buddy through the Buddy CLI, the MCP server and the Buddy skill, and Buddy can run agents for you - as a pipeline action or inside an isolated sandbox.
Which path you take depends on where the work happens. If you code locally and want the agent to handle deployments and pipelines on request, install the skill and point the agent at the MCP server. If you want agent work to happen automatically on events - a pull request opened, an email forwarded, a new Sentry issue - run the agent as a pipeline action. If you want the agent to have a clean, throwaway machine with your repository fetched in, run it in a sandbox.
Everything here works with the agents documented in this section, so you can start with the one you already use and swap later without changing the Buddy side.
In this section
| Page | What it covers |
|---|---|
| Agents | Per-agent guides for connecting Claude Code, Codex, Cursor, Copilot, Antigravity, Devin, Kiro, OpenCode, Grok Build and Pi to Buddy. |
| MCP | Connect an agent to Buddy's remote MCP server so it can run pipelines, manage deployments, sandboxes, packages and domains. |
| Skills | Install the Buddy skill so the agent knows how to link projects, deploy, run pipelines and read logs on its own. |
| Run agents in a sandbox | Spin up an isolated cloud environment, fetch your repository, connect over SSH and run the agent there. |
| Run agents in a pipeline | Add an agent as a build action and automate agent work on pipeline events. |
| Claude Managed Agents | Self-host Claude Managed Agents on Buddy Sandboxes, with an orchestrator and one worker sandbox per session. |
Agents
Guides for connecting AI coding agents to Buddy - deploy projects, run pipelines, and debug builds from the agent of your choice.
MCP
Connect AI agents to Buddy over the Model Context Protocol (MCP) to run pipelines, manage deployments, sandboxes, packages, domains, and more.
Skills
Install the Buddy skill so your AI coding agent knows how to link projects, deploy, run pipelines, and read logs with Buddy.
Run agents in a sandbox
Spin up an isolated Buddy sandbox, fetch your repo, connect over SSH, and run your AI coding agent in a clean cloud environment.
Run agents in a pipeline
Run AI coding agents as a pipeline action to automate agent work on events - PR reviews, email-driven tasks, Sentry issues, and more.
Claude Managed Agents
Self-host Claude Managed Agents on Buddy Sandboxes - Anthropic owns the agent loop, Buddy runs the execution layer with an orchestrator and one isolated worker sandbox per session.