Your KV Cache Can Spot Attacks on AI Agents: Intrinsec AI's Approach
AI Engineer · youtube · 2026-11-09
Intrinsec AI founder Ari Azarafrooz shows how the KV cache your inference engine already writes can double as a security signal for AI agents. Instead of judge models or per-failure classifiers, his system learns a baseline from normal traffic and flags anomalous KV pages, catching issues like a hidden spoken instruction in a voicemail where no text exists to filter. He covers integration points with vLLM, SGLang, LMCache, llm-d and NVIDIA Dynamo, plus two production hard parts: constantly changing models and KV data too large to move in real time.
More from coding & agent
- Getting Omarchy running on Snapdragon Surface Pro: custom kernel plus SSH tunnel for OpenClaw — vincent_koc · 2026-11-09
- Resurfacing in Obsidian: Random Notes, Review Queues and Related Notes — dSebastien · 2026-11-09
- One prompt edit made an agent burn $133 on 21 rogue calls — and the 3 gates that fixed it — Worth_Wealth_6811 · 2026-11-09
- Solo dev adds read-only MCP connector so Claude and ChatGPT can query your food diary — Content-Two5729 · 2026-11-09
- Suspected DPRK npm supply chain attack: archetype-index-sdk ships obfuscated loader with credential harvesting — cyb3rops · 2026-11-09
- OpenBrowser: open-source AI agent browsing toolkit hits 9.6k GitHub stars in days — thisdudelikesAI · 2026-11-09