Review Multi-Agent System Designs and Implementations

This skill performs review and analysis of multi-agent orchestration designs or implementations. It examines architecture, configuration, and code without taking any action that would modify systems, launch processes, or alter production environments.

Use this skill when you need to evaluate a multi-agent system design before implementation, audit an existing orchestration setup for potential issues, or get feedback on architecture decisions without risk of accidental execution. It is ideal for code review sessions, design validation, and pre-deployment checks where you want analysis only.

Do NOT use this skill when you actually need to implement changes, deploy fixes, launch workers, mutate queues, cancel runs, or change production configuration. Those actions require separate, explicit requests after the review is complete.

You provide the design documents, implementation code, configuration files, or architecture diagrams for your multi-agent orchestration system. The skill returns a review with observations, potential issues, and recommendations without executing any of the reviewed operations.

This skill serves engineering teams working with multi-agent systems who need safe review capabilities. It is particularly valuable for DevOps engineers and architects who want to validate orchestration logic before deployment, and for developers who need a second opinion on agent coordination patterns without risking accidental execution of production commands.