parallel-dispatch

Parallel dispatch orchestrates concurrent execution of multiple agents, launching them simultaneously and collecting their results together. This speeds up reviews like security assessments, architectural evaluations, and requirements analysis by distributing work across agents instead of running them sequentially.

Yes. Parallel dispatch loads agent configurations and injects agent-specific prompts, allowing you to run specialized agents—such as a security architect, test architect, and requirements analyst—on the same artifact with tailored instructions for each role.

Parallel dispatch enforces timeouts on concurrent tasks to ensure timely results. Each agent's execution is bounded, preventing slow agents from blocking the entire review and allowing you to collect partial results within a predictable window.

Parallel dispatch returns per-agent outputs and execution metadata in structured form, consolidating results from all concurrent agents. The output is suitable for downstream processing, analysis, or export.

Parallel dispatch consolidates per-agent outputs into a single result set, enabling you to combine perspectives from different specialized agents—such as security, testing, and requirements experts—into comparable, aggregated findings for comprehensive reviews.