Documentation
Everything you need to red team AI with Basilisk.
Overview
Introduction to Basilisk — the open-source AI/LLM red teaming framework with genetic prompt evolution.
Read →Getting Started
Install Basilisk and run your first AI security scan in under 5 minutes.
Read →Attack Modules
Complete reference for all 29 attack modules across 8 categories, mapped to OWASP LLM Top 10.
Read →Evolution Engine
Deep dive into Smart Prompt Evolution (SPE-NL) — the genetic algorithm that evolves attack payloads.
Read →CLI Reference
Complete reference for all Basilisk CLI commands, flags, and options.
Read →Reconnaissance
5 reconnaissance modules for fingerprinting AI models, profiling guardrails, and discovering capabilities.
Read →Reporting
Generate HTML, SARIF, JSON, Markdown, and PDF reports from Basilisk scans.
Read →Provider Configuration
Configure Basilisk to test any LLM via OpenAI, Anthropic, Google, Azure, Ollama, or custom endpoints.
Read →Architecture
System design overview of Basilisk's core engine, provider adapters, evolution engine, and native extensions.
Read →