domains / families / agent-tooling / parallel-task-spark

parallel-task-spark

separate

Only to be triggered by explicit /parallel-task-spark commands.

When to use

• Parsing markdown plan files
• Delegating tasks to parallel Sparky subagents
• Managing task dependencies and waves
• Validating task completion with TDD or alternative evidence
• Updating plan documents with logs
• Committing task work locally

When not to use

• Triggered by implicit commands
• Using non-Sparky agent roles
• Pushing commits to remote repositories

Skill page

Source