Dossier
open-source harness
Coverage of open-source harness in the Nexus archive.
- Anthropic’s Alarming Mythos Findings Replicated With Off-the-Shelf AI, Researchers Say
Security researchers replicated Anthropic's Mythos vulnerability findings using GPT-5.4 and Claude Opus 4.6 in an open-source setup, achieving results for under $30 per scan. The findings highlight the accessibility of AI tools for security analysis.