code quality compass
also via · Download macOS app · v1.8.1 changelog
What is quodeq
Hand your repo to a senior engineer and ask for a full audit. That's quodeq, except it takes minutes instead of weeks.
An AI agent reads your code file by file, follows the imports, and tells you what's solid, what's broken, and how to fix it. Every score backed by ISO 25010. Every finding mapped to a CWE. Not a linter. It actually reads your code.
One real project, from first scan to fix plans. More videos
What it finds
query = f"SELECT * FROM users WHERE id = {user_id}"
credentials = {"user": "admin", "pass": "secret123"}
except: pass
cursor.execute("SELECT * FROM users WHERE id = ?", (user_id,))
This is real output. The offending line, the reason, the CWE, and a fix plan you can paste straight into your agent. Compliant findings count too, they cite the CWE the code defends against. Everything lands as JSON on your machine.
Point it at any repo. The agent explores with read-only tools, follows the imports, and builds real context. Any language, any framework.
Findings map to ISO 25010 dimensions and roll up through the Q² formula. Good code raises your score. It's not just a penalty machine.
Every violation shows the exact line, the reason, and a fix plan. Paste it into your agent or IDE and get on with your day.
Straight answers
Nothing. MIT licensed. Run Ollama and the AI is free and fully offline too. Or bring the Claude, Codex, or Gemini setup you already use.
Any. The agent reads code the way an engineer does, so there's no parser to limit it. Python, TypeScript, Go, Java, whatever you ship.
Nowhere. Evaluation runs on your machine and results live in ~/.quodeq as JSON. With Ollama, nothing ever leaves.
ISO 25010 and CWE built in. Add your own standards for yourself, your team, or your company.
Needs Python 3.12+, or grab the macOS app.
Works with