Explore hands-on experiences, communities, and interactive spaces.




Destrier is launching its first product at OB1: an AI-first cyber range for evaluating autonomous security agents against controlled, realistic targets. Each run takes place in an isolated, disposable environment with fixed scope, approved model access, and time and cost limits. Destrier records the agent's activity, verifies captured objectives using platform evidence, and produces a replayable trace and scorecard.
Across the two-day event, Destrier will run multiple guided mini-CTF sessions, each lasting approximately 90 minutes and accommodating up to six participants. Workstations and prepared harnesses will be provided. Participants configure how their agents should operate before launching them against the same challenge box under identical conditions. Once the round begins, the agents run autonomously without human steering, while participants and spectators follow their attack paths, verified captures, and results live.
These sessions are part of Destrier's competition arm, where we are building new ways for autonomous agents to be evaluated. Come talk to us after a round, whether you played or watched. We'll unpack what happened, introduce you to Destrier's autonomous security evaluation product, and share where we're taking it next.
Difficulty level:
Suitable for all experience levels. Participants will receive a guided setup on the day and configure a prepared harness through high-level settings, including but not limited to prompts, model behaviour, reconnaissance, recovery strategy, context management, and budget awareness. No prior CTF or harness-development experience is required.