Python Developer · AI & Security Tools · Open-source Enthusiast
Evidence-driven AI code audit for Python. Every verdict ships with a machine-checkable taint path (source → hops → sink) — static analysis proves the dataflow, the LLM judges it with receipts, and nothing is ever silently hidden.
pip install aicaudit
aicaudit scan ./src --ai- 🔍 Taint engine — proves constants safe, kills classic SAST false positives
- 🤖 AI verdicts with receipts — three-state, fail-safe, evidence-grounded prompts
- 📊 Honest benchmarks — 100% precision/recall on the labeled corpus, methodology reproducible, Bandit comparison included
- 🖥️ Offline-first Web UI — taint path stepper, fix diff-preview/apply/rollback
→ Repository · PyPI · Use it as an MCP server in your AI editor · PyPI · Benchmarks · Changelog
- 🔭 I'm working on AI-powered code audit tools and security automation
- 🌱 I'm currently exploring LLM Agents, Security Automation, and Developer Tools
- 👯 I'm looking to collaborate on open-source security/AI projects
- 💬 Ask me about Python, AI tools, Security, Code Quality
- ⚡ Fun fact: I build tools that make developers' lives easier
| Category | Technologies |
|---|---|
| Languages | Python, Shell, JavaScript |
| AI/ML | LLM APIs, OpenAI, RAG, Agents |
| Security | Penetration Testing, Code Auditing, Vulnerability Scanning |
| DevOps | Docker, Git, CI/CD, GitHub Actions |
| Backend | FastAPI, Flask, REST APIs |
⭐ If you find my projects useful, please consider giving them a star!