Pinned Loading

1. LLM-Vulnerability-Scanner LLM-Vulnerability-Scanner Public

   Serverless Automated Red Teaming Engine for AWS Bedrock. Scans Llama 3, Claude 3, and custom fine-tuned models for vulnerabilities without provisioned infrastructure.

   Python 2

2. RedTeam-Agent RedTeam-Agent Public

   An autonomous, multi-model Red Teaming engine that pits high-intelligence "Attacker" agents against "Victim" models to discover safety vulnerabilities.

   Python 1

3. AISS AISS Public

   AI Security Screener - Comprehensive security testing tool for AI agents

   Python

4. ImageWarden ImageWarden Public

   A multimodal AI security pipeline for detecting visual prompt injections.

   Python

5. Mirage-HoneyPot Mirage-HoneyPot Public

   Agent-aware deception engine. Mirage shifts the cost of attack onto LLM-driven scrapers by exploiting token budgets and context limits. Dual-component: Trap & Radar.

   Python 1

6. Mirage-Sight Mirage-Sight Public

   Automated threat intelligence repository for the Mirage-HoneyPot ecosystem. Aggregates and structures telemetry from adversarial AI agents into daily/monthly feeds. Features automated S3 ingestion,…