architecture-parity-review

To audit Claude Code transcripts for ARCHITECTURE.md parity, you cross-reference session transcripts against the documented architecture to surface drift and misalignments. This process enforces cross-reference checks and generates a synthesized parity report with documented gaps and recommended updates.

Architecture parity ensures ARCHITECTURE.md stays in sync with decisions documented during feature work. You need to audit project transcripts during major feature rollouts, periodic audits, and new-team onboarding to align implementation with governance.

You generate an architecture parity report by running parity analysis on project transcripts, which spawns parallel agents to audit decisions across sessions and projects. The analysis outputs an ARCHITECTURE_PARITY_REPORT.md file detailing documented gaps and recommended updates.

Yes, you can spawn parallel agents to audit architecture decisions across multiple sessions and projects. This parallel-agent analysis verifies ARCHITECTURE.md parity against implemented decisions and synthesizes findings into a comprehensive parity report.

The architecture parity audit is currently project-specific to the Flow budget tracking app. It enforces cross-reference checks and parallel-agent analysis tailored to this context to surface drift and misalignments effectively.

ARCHITECTURE.md drifts from implemented decisions when session transcripts document changes that are not reflected in the architecture file. Cross-referencing transcripts against ARCHITECTURE.md surfaces this drift and misalignment for governance alignment.