By name: use-task-dag-for-coordination description: When coordinating multiple tasks across multiple agents, always use DECLARE_TASKS to create a task DAG instead of manual tracking. Applies whenever a project lead has 3 or more tasks to coordinate.
Always Use Task DAG for Multi-Task Coordination
When you are a Project Lead coordinating multiple tasks across multiple agents, always use DECLARE_TASKS to create a task DAG at the start of the project.
Why
- Manual tracking via TODO lists is error-prone and doesn't auto-schedule
- The DAG system handles dependency ordering, parallel execution, and status tracking automatically
- TASK_STATUS gives you real-time visibility into what's done, what's blocked, and what's ready
- Other agents and the system can query DAG state for coordination
When to Use
- Any project with 3 or more tasks
- Any project with task dependencies (B depends on A finishing first)
- Any project with multiple agents working in parallel
How
- Break the work into discrete tasks with clear IDs
- Identify dependencies between tasks
- Use DECLARE_TASKS with the full task list and dependency graph
- Use TASK_STATUS to monitor progress
- Use QUERY_TASKS to inspect state when debugging
Example
Instead of manually delegating and tracking:
DECLARE_TASKS with tasks:
- {id: 'design', role: 'architect', description: '...'}
- {id: 'implement', role: 'developer', dependsOn: ['design']}
- {id: 'review', role: 'code-reviewer', dependsOn: ['implement']}
- {id: 'test', role: 'qa-tester', dependsOn: ['review']}
The system will auto-start each task when its dependencies complete.