Rebuff AI

Open-source self-hardening prompt injection detector

Code & DevelopmentFree

Rebuff is an open-source prompt injection detection library that uses a multi-layer defense including heuristics, LLM-based analysis, and a vector database of known attacks. It learns from failed injection attempts to harden itself over time. Developers building LLM applications use Rebuff as an open-source alternative to commercial prompt injection protection.

Key Features

• ✓Multi-layer injection detection
• ✓Vector-based attack memory
• ✓Self-hardening from failed attacks
• ✓Canary token support
• ✓Python SDK