Meet your AI team: six agents that ship code

A single general-purpose assistant is a generalist by definition. Real software gets built by people in roles — someone designs, someone implements, someone tests, someone reviews. Swork Team mirrors that with a team of specialized agents.

The roster

• Team Lead — coordinates work across the other agents.
• Product Manager — turns goals into concrete tasks.
• Architect — designs structure and interfaces.
• Engineer — implements and edits code.
• Tester — writes and runs tests.
• Code Reviewer — reviews diffs for issues before they land.

Your models, your rules

Each agent can run on Claude, GPT, Gemini, or a local model — pick per agent or set a default. On a paid plan, agents use managed credits; on the free plan, you bring your own provider keys.

Every agent has configurable permissions for reading files, writing files, and running commands. By default, anything that changes your code surfaces a diff or a confirmation first, so you stay in control.

Agents propose; you decide. Nothing touches your codebase without a review step you can keep or remove.

— The Swork Team