
PurpleLlama
Set of tools to assess and improve LLM security.

Set of tools to assess and improve LLM security.

A prompt injection in a code‑review bot that executes AI‑generated fixes in a sandbox. The sandbox uses a blacklist to prevent dangerous commands,…

Two-stage prompt-injection and jailbreak detector: regex gates plus a quantised DeBERTa-v3 ONNX classifier, with image, document, and audio support.…

The Security Toolkit for LLM Interactions

Anti-LLM obfuscation via finger counting

PoC for CVE-2026-71554 - h2 duplicate Host header request smuggling primitive (fixed in 4.4.1)

A Python toolbox to create adversarial examples that fool neural networks in PyTorch, TensorFlow, and JAX

MAAD Attack Framework - An attack tool for simple, fast & effective security testing of M365 & Entra ID (Azure AD).

Canary Hunter aims to be a quick PowerShell script to check for Common Canaries in various formats generated for free on canarytokens.org

A security scanner for your LLM agentic workflows