review

Pre-landing code review identifies security flaws like SQL injection and shell injection by analyzing git diffs against the base branch. It uses specialist subagents to scan for critical issues before code reaches production, ensuring zero-defect shipping standards.

Two-pass diff analysis is a pre-landing pull request review methodology that first scans git diffs for critical issues like SQL safety and race conditions, then identifies informational issues like async mixing and dead code. It delegates to parallel specialists for testing, security, and performance concerns.

To automate fixing mechanical code issues like N+1 queries in a pull request, apply fix-first review heuristics. This approach auto-fixes mechanical issues including magic numbers while escalating security and design decisions to human judgment during the diff analysis process.

Yes, you can use Greptile integration to triage automated review comments on a pull request. It fetches comments from Greptile, classifies them as valid, false positives, or already fixed, and replies with evidence-based responses during the code review process.

In a pre-landing code review, you should look for security vulnerabilities including SQL injection, shell injection, race conditions, and LLM trust boundary violations. Specialist subagents analyze the git diff to detect these critical flaws and enum completeness gaps before merging.

Pre-landing diff analysis works without external dependencies by using specialist subagents to scan git diffs internally. However, Greptile integration is available as an optional component for fetching and triaging automated review comments on pull requests.