Agent OPFOR: Open-Source AI Agent Red Teaming Tool
grajmanu · reddit · 2026-07-08
Agent OPFOR is an open-source adversarial simulation tool for AI agents. Named after the military 'OPFOR' (Opposing Force) concept, it conducts red teaming on agents like a real adversary rather than relying on static evals. The attack surface covers multi-turn prompt injection and jailbreaks, system prompt extraction, tool abuse and BOLA/BFLA, MCP endpoint attacks (tool description injection, secret leakage, privilege escalation, SSRF), memory poisoning, goal hijacking, and EU AI Act bias testing. Its opfor hunt mode allows a command agent to autonomously plan an attack campaign given specific endpoints and targets, with support for viewing the attack tree in real time.
Related event: Agent OPFOR: Open-Source Red Teaming Tool for AI Agents(3 posts)→
More from Safety
- Ex-DOGE cyber researchers breach Anthropic's systems with AI agents in under 12 hours — GarrisonLovely · 2026-10-27
- Alignment Hub launches with 120+ documented AI misbehavior incidents and a reporting template — Thom_Wolf · 2026-10-27
- Alignment Hub launches to track 37 misalignment incidents with disclosure stats — eliebakouch · 2026-10-27
- US senator slams AI industry for proposing 50-death threshold for 'catastrophic incident' — peterwildeford · 2026-10-27
- 125 poisoned samples backdoored Qwen2.5-7B — and open-weight supply chain risks debated — lxrjl · 2026-10-27
- Dev training a guard model to block prompt injection, tool and memory poisoning — gnukeith · 2026-10-27