opencode

To delegate coding tasks and refactoring, route your implementation work to the OpenCode CLI agent, which processes feature development and bug fixes autonomously. You simply specify the repository, goal, and whether you want a one-shot run or an interactive session.

Yes, you can run parallel development tasks in isolated workdirs or worktrees to avoid collisions. This allows separate autonomous tasks to execute simultaneously without interfering with your main terminal workflow.

You monitor long-running coding agent sessions by starting interactive TUI-based terminal sessions through the OpenCode CLI. This allows you to track progress, monitor agent status, and resume work when needed.

To use an autonomous coding agent for CLI automation, you need the OpenCode CLI installed, authenticated provider access, and an interactive terminal that supports TUI-based sessions for isolated workdirs and progress monitoring.

Yes, you can use a coding agent to review pull requests and find security risks. The agent inspects PRs to surface bugs, test gaps, security vulnerabilities, and style issues automatically.

Autonomous development supports long-running debugging sessions by offloading failing test debugging to an external coding worker. You can start an interactive session, monitor progress, and resume the debugging work whenever needed.