Fresh Eye

Code Review Sanity Checks

Overview

Self-review is a critical step in the software development lifecycle. This skill automates the analysis of staged and unstaged code modifications (git diff) to identify common errors, logic gaps, debug statements, and coverage omissions before code is committed. It prints findings directly to standard output (stdout) and sets exit codes to gate commits.

Automated Execution

To run the automated code review script, execute it from the root of the workspace:

./fresh-eye/fresh-eye.sh

Exit Codes

• 0: Success. No blocker or high-priority issues were detected.
• 1: Failure. Blocker or high-priority issues were detected, indicating the commit should be halted.

Output Structure

The script prints a structured review directly to standard output using the following format:

• Scope of Changes: Summary of modified files.
• Blocker / High Priority Issues: Critical logic errors, debug code (e.g., console.log, print), or safety risks. If present, these trigger a non-zero exit code.
• Minor / Refactoring Suggestions: Code smell detections, style improvements, or optimization recommendations.
• Test Coverage Audit: Flags exports or modules missing associated test files.

Verification Checklist

Prior to completing the review, verify that:

1. Exit Code Correct: The script exits with status 1 if and only if high-priority blockers are present.
2. Stdout Formatted: Output is clean and readable in the terminal.