#110 · Primary category: Cybersecurity & Decryption Tools

stride-gpt

ai-security application-security attack-trees cybersecurity dread generative-ai llm mcp mitre-attack security-tools streamlit stride threat-modeling

An AI-powered threat modeling tool that leverages OpenAI's GPT models to generate threat models for a given application based on the STRIDE methodology.

Project last updated:08/28/26

GitHub Stars

1.1K

Forks

321

Contributors

10

License

MIT

Why we included this project

STRIDE is the classic way to work through a system's threats, but doing it properly by hand is slow, which is where this tool earns its keep. Give it the basics about your application, like what type it is, how users authenticate, and whether it sits on the internet or handles sensitive data, and it produces threat models and attack trees instead of leaving you staring at a blank page. Beyond single prompts, an agentic mode can work through a whole codebase on its own, and there's specialized analysis for agentic AI and GenAI stacks aligned with the OWASP ASI and LLM Top 10 lists. Output is flexible too: terminal users get a full CLI with an interactive REPL, while a Streamlit web UI covers everyone else, and each threat can carry MITRE ATT&CK and ATLAS technique identifiers so findings drop into existing security reporting without extra wrangling.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category