heavy3-aiheavy3-aiOfficialยท1 Agent Skills Included

code-audit

Multi-model code reviews for commits, plans, and pull requests

Reviews code changes, implementation plans, and GitHub pull requests using single or three-model consensus checks. Catches bugs, security flaws, and performance issues that a single reviewer or model would miss. Estimates review cost upfront, splits oversized changes into modules, and works with free or paid models via your own API key.
npx skills add heavy3-ai/code-audit --all -g -y
Available:

Explains to the AI agent how the /h3 review skill is structured, how context flows from git diffs into the review scripts, and how to run and test each review mode.

All Skills in This Repository (1)

Pure Emerald Level Indicators

Frequently Asked Questions

FAQPage Schema
How to install Heavy3 Code Audit?โ–ผ

Run `npx skills add heavy3-ai/code-audit --all -g -y` in your terminal to install the skill globally, then add your OpenRouter API key.

What does the council review mode do?โ–ผ

Council mode runs three models in parallel (GPT, Gemini, and Grok), each focused on correctness, performance, or security, then merges findings into one comparison table.

Can I review code for free?โ–ผ

Yes. Use the --free flag to route reviews through rotating free models on OpenRouter at zero cost.

Does it work with Claude Code and Cursor?โ–ผ

Yes. The skill follows the standard SKILL.md format and runs in Claude Code, Cursor, Codex CLI, and Gemini CLI.

Can it review large pull requests?โ–ผ

Yes. It detects oversized changes, splits them into module-by-module reviews, and shows a cost estimate before running.

Related Repositories in Software Engineering

View All in Software Engineeringโ†’