TLDR Tech / InfoSec
RAPTOR (GitHub Repo)
RAPTOR is an autonomous security research framework built on Claude Code that integrates static analysis, binary analysis, LLM-driven vulnerability validation, exploit generation, and patch writing into a single workflow. It ships over 200 custom Semgrep, Coccinelle, and CodeQL rules, a six-stage exploitability validation pipeline that filters scanner noise before a finding is called real, and optional Z3 pre-screening that drops provably unreachable dataflow paths. Subprocesses that handle untrusted repositories are sandboxed using Linux namespaces, Landlock, and seccomp, and the analysis layer accepts Anthropic, OpenAI, Gemini, Mistral, or local Ollama models with a per-run cost cap.
Read the full story