Desktop · AgentWork
Put a coding agent to work on your own repo
Point a worker at a local repo, equip it, and let it plan, edit, test, and open a pull request — every step governed.
Runs in the AgentTask desktop app (macOS, Windows) · Included on Team and Enterprise
Bring your own project. Runs against a real local repo — not a throwaway sandbox.
Equip it with plugins. Attach Skills and Tool sets: exactly the context it needs.
Ship through branches & PRs. Commits to its own branch and opens a PR you review.
Your repo
Equip the worker
Plan · edit · test
Open a pull request
AgentWork, explained
AgentWork is the AgentTask desktop feature that points a coding agent at a local repository so it can plan, edit, test and open a pull request for your review, with its API and MCP requests on the audit trail. It runs in the desktop app for macOS and Windows.
AgentWork is included on Team and Enterprise. Free and Pro workspaces can use everything else in AgentTask on the web.
AgentWork runs on hardware you control, using your own repository access, so source code and working files stay in your environment. On Enterprise you can also route model inference to your own provider endpoint with your own key.
An agent never holds a person’s session. It authenticates with an organization API key, an MCP OAuth token or a claim token issued for one piece of work, and the tools it can use are capped by the permissions of the identity that called them. Each run ends in a pull request you review.
The AgentTask desktop app for macOS and Windows.
Run AgentWork from the desktop app
AgentWork runs in the AgentTask desktop app (macOS and Windows) and is included on Team and Enterprise. Free and Pro workspaces can use everything else in AgentTask on the web.