| name | team |
| description | Activates parallel team execution — N coordinated agents working simultaneously on independent subtasks |
| tags | ["execution-mode","parallel"] |
Activation
This skill activates the team agent for parallel multi-agent execution.
Usage
team <task description>
team N <task description>
team N:agent-type <task description>
team ralph <task description>
Parameters:
- N — number of agents (1-20), auto-sized if omitted
- agent-type — executor (default), debugger, designer, test-engineer
- ralph — modifier: wrap in persistence loop
Trigger Keywords
team, assemble a team, team N, coordinated parallel execution
Plan Discovery
Team checks for existing plans before decomposing:
store_memory key omg:active-plan
.omg/plans/ directory
Dependency & Conflict Management
- Each worker gets explicit file boundaries (no overlap)
- After all workers complete: combined build + test verification
- If conflicts detected: diagnose which worker caused the issue, fix that one
Quality Contract
- Parallel, not sequential — never serialize independent work
- Combined verification after all workers complete
- Progress tracked in
.omg/qa-logs/team-log.md
Example
copilot -i "team 3: fix TypeScript errors across src/"