← sources

Has anyone found a reliable way to scan agent skills for security risks without drowning in false positives? I’m looking for something that can distinguish genuinely unsafe behavior from legitimate tool use, rather than flagging anything that looks suspicious in isolation.

source: reddit (reddit:AI_Agents)published 2026-09-27 02:14ingested 2026-09-28 02:36

retained text

We've spent way too much time looking into ways to check agent skills before using them and I still don't feel great about any of it. AI review can get prompt injected by the thing it's reviewing, scanners flag stuff the skill legitimately needs to do, and so many findings end up needing a manual review anyway. Every approach we've looked at seems to have its own version of this problem.

I'm not expecting perfect security, just something that saves us some work without constantly flagging harmless stuff. Has anyone actually gotten this to a point where they're happy with it? Would love to hear what's working for people because we're going in circles a bit.

document fetched 2026-09-28 00:19 · canonical https://reddit.com/r/AI_Agents/comments/1wr86p2/has_anyone_found_a_reliable_way_to_scan_agent