multi-agent-red-team-planning

Multi-agent red-team planning is a structured workflow that orchestrates multiple sub-agents to critique repository changes across security, reliability, performance, and code quality, producing actionable implementation plans and formal artifacts.

You execute a staged multi-agent planning workflow that includes sub-agent coordination, critique rounds across domains like security and performance, external research synthesis, and formal delivery of implementation plans and roadmap updates.

Yes, the workflow applies to planning repository changes, feature branches, security evaluations, and risk scenarios across multiple sub-teams, coordinating agents to generate red-team reports and research summaries.

The process generates formal artifacts including a final implementation plan, red-team reports covering security and reliability, research summaries from external synthesis, and roadmap updates for ongoing project governance.

The red-team critiques cover six dimensions: Security, Reliability, Performance, UX/DX, Code Quality, and QIG Purity, ensuring comprehensive risk assessment before finalizing the implementation plan.

No dependencies are required. The workflow operates as a standalone structured process that orchestrates sub-agent coordination, critique rounds, and research synthesis to deliver formal planning artifacts.