Agen

Meet Agen

Agen runs AI coding agents in cloud workspaces connected to Git repositories. Users describe a task, and an agent explores the code, prepares changes on a separate branch, runs services, and works through pipeline failures. Live previews let a reviewer inspect the running application alongside the proposed code changes before deciding whether to merge.

A session can span related frontend, backend, and infrastructure repositories. Scheduled agents support recurring maintenance with budget controls, while parallel sessions provide separate places for different tasks. The browser interface allows team members to assign work without preparing a local development environment, with a technical reviewer retaining responsibility for the final changes.