Code Review

Systematic examination of code changes to catch defects, enforce standards, and share knowledge across the team.

Context

Code review is the last quality gate before code enters the shared codebase. It catches bugs that automated tests miss, prevents security vulnerabilities from reaching production, distributes system knowledge across the team, and maintains architectural consistency. Effective reviews balance thoroughness with turnaround speed -- a review that takes three days damages velocity more than the bugs it catches.

Inputs

• source-code: The diff or changeset under review, with sufficient context to understand the change.
• test-suite: Accompanying tests that validate the change. Verify they exist and are meaningful.
• task-list: The reviewed implementation slice or task context that defines what was actually supposed to land now.
• api-contract: The contract or externally visible behavior that the changeset must preserve or implement.
• architecture-doc: System design context to verify the change aligns with architectural decisions.
• intake-brief: Must include quality_target_context with runtime_context, exposure_profile, production_target, non_targets, and evidence_refs.

In a lifecycle-aware system, review should not silently approve code that closes unresolved upstream questions by accident. Brownfield coexistence, unsupported release-1 flows, and contract boundaries are review concerns, not "later" concerns.