parallel-agents

Parallel task dispatch assigns isolated agents to each independent problem, enforcing self-contained contexts to prevent cross-contamination while executing concurrent investigations.

Orchestrating dedicated agents for separate problem domains allows concurrent investigation of unrelated failures, reducing total turnaround time by aggregating parallel results without shared state.

Use parallel agents when you face 3 or more unrelated failures or problems across independent domains that can be investigated concurrently without shared state or data leakage.

Provide three independent issues and instruct the system to assign them to separate agents, which creates self-contained prompts per agent to maintain isolation and predictability during parallel dispatch.

No, parallel agents enforce isolated contexts per agent to prevent cross-talk, making this approach unsuitable for problems requiring shared state or dependent data across concurrent tasks.