code_review

To analyze technical debt and system architecture, you provide a task and relevant project files to receive structured, architecture-level reasoning. This process evaluates code structure and design patterns to guide strategic decisions without requiring implementation details.

Planning system improvements for a failing module involves providing the specific files and a task description to an architect-style analysis. This generates implementation-oriented plans and sequencing steps to resolve the regression and improve the system design.

Root cause analysis for debugging a codebase regression works by evaluating selected files and optional git diffs to identify why behavior breaks. It performs architecture-level reasoning to recommend targeted fixes and strategic next steps for the failing module.

Yes, you can include git diffs and commit ranges alongside your task and relevant files to evaluate code structure. This provides the context needed to produce strategic architectural guidance and identify technical debt introduced by recent changes.

Yes, providing relevant files is required to get architectural guidance for your code. The analysis depends on evaluating the selected files and your defined task to return structured results with strategic recommendations.

Architectural analysis requires an architect subagent to evaluate code patterns because it provides deep, strategic reasoning over your selected files. This mechanism ensures the system design and technical debt are evaluated at an architecture level rather than focusing on line-by-line implementation.