code-review

To perform a security-oriented code review for LoongCollector PRs, you can use this skill to fetch changes, build context, and map incremental reviews to deliver architecture-first findings. It analyzes branches and PRs to identify defects and design inconsistencies.

Architecture-first code review evaluates module boundaries and dependencies alongside security risks. For incremental changes, it maps patch and hunk analysis across multiple files to provide snapshot-driven decisions and track review conclusions.

Yes, you need python3, git, and gh installed to review LoongCollector branches. These dependencies are required to fetch repository changes, map incremental hunks, and report deterministic code review findings.

Yes, you can audit module boundaries and dependencies for large commits. The skill performs end-to-end context building and impact analysis to track architectural alignment and security vulnerabilities before merging.

The best way to track review conclusions across multiple files in GitHub is using snapshot-driven decision support. This approach applies deterministic patch and hunk analysis to map findings consistently across reviewers.