Full talk: pattern lists stop under 25% of prompt injections, guard models catch most
colinmcnamara · x · 2026-10-15
colinmcnamara released the full video of his AustinLangChain talk. Core experiment: he tried to sneak prompt injections past his own router — his current pattern list caught under a quarter of them, while the guard model he measured caught most. Takeaway: pattern matching isn't enough; wire a guard model into the pipeline. Full walkthrough in the video.
More from coding & agent
- Open-Agent-Email: free open-source alternative to AgentMail for AI agents — matchaman11 · 2026-10-15
- AWS Gen AI Weekly Redesign: One-Page Exec Summary, AgentCore Memory Goes Usage-Based — DigitalColmer · 2026-10-15
- Nous Research Launches 50+ Connectors for Hermes Agent, No Account Required — NousResearch · 2026-10-15
- Dev group chat: skip runtime LLM web search, index products via API instead — vxnuaj · 2026-10-15
- Tev-Gym: 1,170 dense-reward RL environments push long-horizon agentic training — RulinShao · 2026-10-15
- Microsoft's AutoEnvScaling lets agents build their own RL environments, lifting Qwen3 Terminal-Bench from 41.1 to 53.3 — omarsar0 · 2026-10-15