A Playbook is a pre-built mission template that defines how your AI team should work together to achieve a goal. Think of it as an operational manual for your AI department.
What a Playbook defines
- Roles — Which specialist roles are needed (e.g., Coordinator, Architect, Developer, QA).
- Workflow stages — The sequence of work: requirements → architecture → implementation → testing → review.
- Coordination rules — How agents hand off work, review each other's output, and handle failures.
- Required context — What information the Playbook needs upfront (repository URL, design specs, etc.).
Available Playbooks
| Playbook | Best for | Roles involved |
|---|---|---|
| SDLC Website | Full-stack web development | Coordinator, Architect, Developer, QA |
| Generic | Flexible, single-agent tasks | Any available agent |
More Playbooks are being added regularly. You can view and manage available Playbooks from Settings → Playbooks.
How Playbooks drive missions
When you launch a mission:
- The Playbook's root instruction is assigned to a Coordinator agent.
- The Coordinator reads your mission objective and the Playbook's workflow definition.
- It decomposes the goal into a sequence of tasks and dispatches them to specialist agents.
- Each specialist follows the Playbook's stage-specific instructions.
- Built-in quality gates ensure outputs are reviewed before moving to the next stage.
Tip: You don't need to understand the internals of a Playbook to use it. Just select the right one for your goal, describe your objective, and let the system handle orchestration.