SkillsApr 5, 2026·3 min read

RAPTOR — Security Research Agent for Claude Code

Autonomous offensive and defensive security framework built on Claude Code. Performs static analysis, binary fuzzing, vulnerability discovery, exploit generation, and patch development. MIT.

TO
TokRepo精选 · Community
Quick Use

Use it first, then decide how deep to go

This block should tell both the user and the agent what to copy, install, and apply first.

1. Install Claude Code from claude.ai/download 2. Clone and enter the project: ```bash git clone https://github.com/gadievron/raptor.git cd raptor ``` 3. Launch Claude Code: ```bash claude ``` Claude will auto-install dependencies and configure the security toolkit. **Alternative (DevContainer with all tools pre-installed):** ```bash docker build -f .devcontainer/Dockerfile -t raptor-devcontainer:latest . docker run -it raptor-devcontainer:latest ``` ---
Intro
RAPTOR is an autonomous security research framework built on Claude Code that combines traditional security tools (Semgrep, CodeQL, AFL++) with AI-powered analysis. With 1,800+ GitHub stars and MIT license, it turns Claude Code into a security research agent capable of static analysis, binary fuzzing, vulnerability discovery, exploit proof-of-concept generation, and automated patch development. Created by security researchers Gadi Evron, Daniel Cuthbert, Thomas Dullien (Halvar Flake), and others, RAPTOR uses a tiered skill system with 9 expert security personas. Best for: security researchers, penetration testers, and developers who want AI-assisted vulnerability analysis. Works with: Claude Code, Semgrep, CodeQL, AFL++, Playwright. Setup time: under 5 minutes (or instant with DevContainer). ---
## RAPTOR — Framework Architecture ### Command Reference | Command | Purpose | |---------|---------| | `/raptor` | Main security assistant — general queries | | `/scan` | Static analysis with Semgrep/CodeQL + LLM validation | | `/fuzz` | Binary fuzzing with AFL++ and crash analysis | | `/agentic` | Full autonomous security workflow | | `/exploit` | Proof-of-concept exploit generation | | `/patch` | Automated security patch development | | `/understand` | Adversarial code analysis | | `/oss-forensics` | Evidence-backed GitHub repository investigations | ### Tiered Architecture RAPTOR uses a sophisticated loading system to manage context: **Bootstrap (Always Loaded):** - `CLAUDE.md` — Core instructions and security mindset **Tier 1 (Auto-Loaded Contextually):** - Adversarial thinking frameworks - Recovery strategies for failed analyses - Security-first reasoning patterns **Tier 2 (Loaded on Request):** Nine expert security personas modeled after renowned researchers: - Mark Dowd (vulnerability analysis) - Binary exploitation specialists - Web application security experts - Cryptography advisors - And more... **Agents:** - Offensive security specialist with autonomous capabilities - SecOpsAgentKit integration for penetration testing ### Core Capabilities #### Static Analysis (`/scan`) - Runs Semgrep and CodeQL scans - LLM-powered validation to reduce false positives - Prioritizes findings: `Impact × Exploitability / Detection Time` #### Binary Fuzzing (`/fuzz`) - AFL++ fuzzing with intelligent seed generation - Automated crash analysis and triage - Reproducible test case generation #### Vulnerability Analysis - Advanced reasoning to identify exploitable patterns - Cross-reference with known vulnerability databases - Severity scoring and prioritization #### Exploit Generation (`/exploit`) - Proof-of-concept code for confirmed vulnerabilities - Responsible disclosure formatting - Impact demonstration #### Patch Development (`/patch`) - Automated security fix generation - Regression test creation - Code review-ready patches ### Security Note RAPTOR is designed for authorized security testing, defensive security research, and educational purposes. Always ensure you have proper authorization before testing systems. ### FAQ **Q: What is RAPTOR?** A: An autonomous security research framework for Claude Code that performs static analysis, fuzzing, vulnerability discovery, exploit generation, and patch development using AI-powered analysis. **Q: Is RAPTOR free?** A: Yes, MIT license. Some underlying tools (Semgrep, CodeQL) have their own licensing terms. **Q: Who created RAPTOR?** A: Security researchers Gadi Evron, Daniel Cuthbert, Thomas Dullien (Halvar Flake), Michael Bargury, and John Cartwright. ---
🙏

Source & Thanks

> Created by [Gadi Evron](https://github.com/gadievron) and team. Licensed under MIT. > > [raptor](https://github.com/gadievron/raptor) — ⭐ 1,800+ Thank you to the RAPTOR team for building an AI-powered security research framework on Claude Code.

Discussion

Sign in to join the discussion.
No comments yet. Be the first to share your thoughts.

Related Assets

Claude Reflect — Self-Learning for Claude Code

A Claude Code plugin that captures corrections and preferences during development, then syncs learnings to CLAUDE.md automatically. Makes Claude remember across sessions. 870+ stars.

TokRepo精选

n8n-as-code — AI Agent Superpowers for n8n

Give your AI agent n8n superpowers with 537 node schemas, 7,700+ templates, and TypeScript workflow definitions. Works with Claude Code, Cursor, VS Code, and OpenClaw. MIT license.

TokRepo精选

Self-Hosted AI Starter Kit — Local AI with n8n

Docker Compose template by n8n that bootstraps a complete local AI environment with n8n workflow automation, Ollama LLMs, Qdrant vector database, and PostgreSQL. 14,500+ stars.

TokRepo精选
◈Home🔍Search👤Me