OpenClaw

OpenClaw is an ecosystem that contains skills published by community contributors, which are used to evaluate the detection of malicious intent and runtime verification of these skills in open agentic platforms.

Papers using OpenClaw (2)

Red-Teaming Agent Execution Contexts: Open-World Security Evaluation on OpenClaw2026BadSkill: Backdoor Attacks on Agent Skills via Model-in-Skill Poisoning2026