design-review-deep

To conduct a software architecture design review, provide the design document and relevant code context to initiate a deep, multi-agent validation process that analyzes API boundaries and technical documents. The process utilizes fresh-context reviewers to challenge the design from different angles.

Multi-agent validation for technical document analysis is a process where fresh-context reviewer candidates challenge a design from various angles, performing boundary, domain, evolution, and simplicity reviews. It directly validates every candidate in the session to ensure confidence in findings.

Yes, you can customize the scope of a design review by resolving the scope and effective diff based on the conversation, artifact, or file. This ensures the multi-agent validation targets the specific technical document or architecture context you need analyzed.

No, the design review process does not rely on external agents or shellouts for validation. It ensures direct validation by the agent within the session, validating every candidate directly to maintain confidence in the architectural findings.

The best way to get a structured report for software design issues is to use a multi-agent design review process that provides deduplicated issues ranked by severity. This ensures thorough analysis and direct validation of architectural findings without relying on external tools.