# LLMs.txt - AI/ML Models and Security Information
# https://redbolt.ai/llms.txt

# Garak - LLM Security Scanner
# Website: https://redbolt.ai
# Documentation: https://docs.redbolt.ai

# Supported Models and Providers
models:
  - OpenAI GPT-3.5, GPT-4, GPT-4 Turbo
  - Anthropic Claude 2, Claude 3 (Haiku, Sonnet, Opus)
  - Google Gemini Pro, Gemini Ultra
  - Meta Llama 2, Code Llama
  - Mistral 7B, Mixtral 8x7B
  - Cohere Command, Command R
  - Hugging Face Transformers
  - Local models via Ollama, vLLM
  - Azure OpenAI Service
  - AWS Bedrock models
  - Together AI models
  - Replicate models

# Security Testing Categories
vulnerabilities:
  - Prompt injection attacks
  - Data extraction attempts
  - Jailbreaking techniques
  - Bias and fairness issues
  - Hallucination detection
  - Privacy leakage
  - Adversarial inputs
  - Social engineering
  - Code injection
  - Malicious content generation

# Research Areas
research:
  - LLM red teaming methodologies
  - AI safety evaluation frameworks
  - Automated vulnerability discovery
  - Security benchmarking
  - Responsible AI deployment

# Contact
contact: security@redbolt.ai
updated: 2024-01-15