OpenClaw adds Tencent's AI-Infra-Guard to ClawHub's skill security review pipeline
heyneighbor · x · 2026-10-03
OpenClaw has integrated Tencent's AI-Infra-Guard (AIG) into ClawScan, the open-source CLI that powers security review for every skill and plugin uploaded to ClawHub.
- ClawScan runs Tencent AIG and NVIDIA's SkillSpector independently, then an AI judge reviews both scanners' findings alongside the uploaded files for a final assessment
- AIG's skill scanner is an "LLM-driven multi-stage code audit and vulnerability review pipeline" that traces relationships between a skill's instructions and its scripts, dependencies, and data flows, covering nine risk categories including instruction hijacking, memory poisoning, remote payload execution, and insecure dependencies
- The team benchmarked the combined system with Tencent on a fixed 556-case subset of SkillTrustBench, a public benchmark from Tencent and CUHK-Shenzhen, with an ongoing feedback loop
More from coding & agent
- Devs flag GPT's defensive programming for hiding bugs; OpenAI asks for /feedback cases — pvncher · 2026-11-19
- Flutter ships A2UI 1.0: agents respond with real interactive widgets, not chat text — rseroter · 2026-11-19
- Skills vs MCP vs RAG vs Memory: the four context mechanisms AI agents actually need — MaryamMiradi · 2026-11-19
- Anthropic details agent memory: 'dreaming' lets Claude organize its experience — daniel_mac8 · 2026-11-19
- The case for externalized memory as the real fix for continual learning — daniel_mac8 · 2026-11-19
- FT: US tech outages blamed on rush to fix bugs found by AI — nordicinst · 2026-11-19