Automated Attack-Pattern Miner Finds Real Agent Framework Flaws: 700 Paths, 3 Confirmed
_h4xr · reddit · 2026-11-17
A security researcher built a static-analysis pipeline that learns attack patterns in agent frameworks without hand-written rules: build a code graph, propose candidate sources/sinks/guards, filter with a model, score and dynamically confirm. Of 700+ candidate paths, 3 were confirmed and reported to vendors. General lessons: treat model output as untrusted input (sandbox execution), restrict fetch tools from internal network endpoints by default, and keep permissions narrow and per-app. Advocates zero-trust for both user queries and LLM responses.
More from coding & agent
- unslop: a Claude skill that de-slops vibecoded UI hits 17.5k GitHub stars — socialwithaayan · 2026-11-17
- KORMo-Web-11B, a Korean VLM for GUI and web agents, trends on Hugging Face — KORMo-VL · 2026-11-17
- User builds fully automated video creation pipeline with Grok Bot, friend gets hooked — XFreeze · 2026-11-17
- Google DeepMind × MIT work challenges 'more agents = better', seeks scaling principles for agentic systems — gerardsans · 2026-11-17
- Dev builds holiday card maker where Gemini writes p5.js code instead of generating pixels, for consistent style — thisiskp_ · 2026-11-17
- Grok Bot 1.0 ships team sharing and meeting bots, both features requested by users on X weeks earlier — elonmusk · 2026-11-17