dispatching-parallel-agents

Debug multiple independent test failures faster by dispatching parallel agents to investigate each failure domain concurrently, provided the subsystems don't share state or require sequential understanding. This isolates context and resolves unrelated bugs without sequential debugging delays.

Use parallel debugging agents when you have three or more failing test files from different subsystems and can confidently treat each failure as an independent problem without shared state or sequential dependencies.

Scope prompts for parallel agents by defining narrow boundaries per agent, providing self-contained context and expected outputs so each delegated agent receives only the instructions it needs without inheriting broader session history.

Yes, you can run root cause analysis concurrently without context contamination by ensuring each agent receives isolated, self-contained instructions and does not inherit the broader debugging session history, preventing cross-contamination of failure domains.

Integrate results after parallel debugging by reviewing agent summaries, detecting edit conflicts, and validating the integration by running the full test suite to ensure all independent fixes work together correctly.

Parallel agents for test failure isolation cannot be used when failures share state, require sequential understanding, or involve race conditions across subsystems, as these require correlated debugging rather than isolated investigation.