AVE, an open standard ID-ing agentic AI vulnerability classes, crosses 83 records
SelectionBitter6821 · reddit · 2026-09-22
AVE, an open-source standard launched just five months ago, has been listed in awesome-agent-skills-security's Tools & Frameworks section and now covers 83+ records. It assigns stable IDs to behavioral vulnerability classes in agentic AI components — MCP servers, skill files, system prompts — so different scanners flagging the same issue can be reconciled as one finding instead of duplicate alerts.
More from Safety
- ChatGPT allegedly emailed the FBI on its own, joining a week of rogue agent incidents — JanJanJaJa · 2026-09-22
- Coding agent found a stray Gemini key in the repo and burned $40 overnight — Future_AGI · 2026-09-22
- Math isn't as open as claimed: NSA kept differential cryptanalysis secret since 1974 — lpachter · 2026-09-22
- Microsoft internal doc calls generative AI a "doom loop" killing the entire web — DocPNess · 2026-09-22
- MCP design lessons: rules agents should follow go in prompts, rules they must not break go in the schema — volkansuner · 2026-09-22
- AI agent bypasses airline seat paywall to grab $39 exit-row seat for free — RespondOk9407 · 2026-09-22