Reviewing Changes

Workflow

1. Collect the diff

Run the following to capture all changes including untracked files:

1. •git diff HEAD — staged and unstaged changes to tracked files
2. •git ls-files --others --exclude-standard — list untracked files
3. •For each untracked file, run git diff --no-index /dev/null <file> to produce a diff

Combine all output as the review target. If both the diff and untracked file list are empty, inform the user that there are no changes to review.

2. Review the diff

Scan the diff with the following priorities (highest first):

1. •Bugs and logic errors — off-by-one, null/undefined access, wrong comparison operators, missing return values, infinite loops, race conditions
2. •Unintended behavior changes — accidental removal of logic, swapped arguments, changed defaults
3. •Security concerns — hardcoded secrets, injection vulnerabilities, missing input validation at system boundaries
4. •Obvious improvements — dead code introduced in the diff, clearly redundant operations

Do not comment on style, formatting, naming, or documentation unless it directly causes a bug.

3. Report findings

Present findings as a concise list grouped by file. For each finding:

• •State the file and approximate line in the diff
• •Describe the issue in one sentence
• •Suggest a fix if it is straightforward

If no issues are found, say so briefly. Avoid filler or praise — the goal is speed and signal.