pre-flight

Plan review for security and performance risks involves analyzing PLAN.md and REQUIREMENTS.md to identify architectural anti-patterns, authentication gaps, and bottlenecks. Multi-agent pre-flight assessments evaluate component boundaries, data handling, and user-flow completeness prior to execution.

A pre-flight plan review is a multi-agent analysis that surfaces architecture, security, and performance risks in GSD plans before execution. You need it when validating design boundaries, data handling, and user flows to prevent flawed deployments and reduce investigation time later.

Checking REQUIREMENTS.md for edge cases and failure paths requires evaluating spec flow completeness across user journeys. A spec flow analyzer reviews related context to catch potential user-flow issues, ensuring comprehensive coverage of failure paths before coding begins.

Yes, you can assess architecture boundaries and data handling gaps in PLAN.md before writing code. Parallel agent analysis evaluates component boundaries, security gaps, and data handling protocols directly against the plan, catching design flaws early in the development lifecycle.

The best way to catch authentication gaps and performance bottlenecks is running a parallel four-agent review on the new plan phase. This coordinates architecture, security, performance, and spec flow analysis against PLAN.md to surface risks prior to execution.

Deployments fail due to missed user flows and design flaws when plans lack pre-execution review. Running a multi-agent pre-flight assessment on PLAN.md surfaces architectural anti-patterns, security gaps, and incomplete user flows, preventing flawed deployments before coding starts.