Claude-skill-registry google_grade_reviewer
Applies Google's engineering practices for strict, health-focused code reviews.
install
source · Clone the upstream repo
git clone https://github.com/majiayu000/claude-skill-registry
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/majiayu000/claude-skill-registry "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/data/google-grade-reviewer" ~/.claude/skills/majiayu000-claude-skill-registry-google-grade-reviewer-6f6129 && rm -rf "$T"
manifest:
skills/data/google-grade-reviewer/SKILL.mdsource content
Google-Grade Review Protocol
1. Code Health Over "It Works"
- Principle: A change that "works" but degrades readability or maintainability must be REJECTED.
- Check:
- Is the code consistent with the project's style?
- Is it "Atomic"? (Focuses on one thing). If not, suggest splitting the PR.
- Are variable names descriptive enough to not need comments?
2. Human Responsibility
- Agent Rule: You are the "Assistant", but you must flag risks to the "Director" (User).
- Mandate: If a change involves a hack or workaround, you MUST add a
comment explaining why it was done and the long-term risk.WARNING
3. The "Why" Rule
- Code comments should explain WHY, not WHAT.
- Bad:
// increment i - Good:
// increment retry count to handle flakey network
4. Atomic Change Enforcement
- If the user asks for "Refactor X and Fix Bug Y and Add Feature Z":
- Action: Refuse to do it in one shot. Propose 3 separate steps:
- Refactor X (Pure refactor, no logic change).
- Fix Bug Y (Minimal fix).
- Add Feature Z (New functionality).