Code Review

Three-axis review of the diff between HEAD and a fixed point the user supplies:

• Correctness: would this change break something? Bugs, failure modes (what a side effect does when it fails, repeats, or races), security, data, broken contracts, missing tests.
• Standards: does the code conform to how this repo writes code?
• Spec: does the code faithfully implement the originating ticket / PRD / spec?

The axes run separately so they don't pollute each other's context, then this skill aggregates their findings. Prefer parallel sub-agents when the harness supports them; otherwise run the reviews inline in sequence in this session. Either path must finish every axis before aggregating: never leave a review handle running past session end.

Process

1. Pin the fixed point

Whatever the user said is the fixed point: a commit SHA, branch name, tag, main, HEAD~5, etc. If they didn't specify one, ask for it (when invoked from implement, it's the ticket's base commit).

Capture the diff command once: git diff <fixed-point>...HEAD (three-dot, so the comparison is against the merge-base). Also note the list of commits via git log <fixed-point>..HEAD --oneline.

When the caller passes a path scope (as implement does, so parallel sessions sharing one branch don't pollute the range), append it to both: git diff <fixed-point>...HEAD -- <paths> and git log <fixed-point>..HEAD --oneline -- <paths>.