How I Jailbroke an Autonomous AI Agent — Exposing 12 Tools, 23 API Actions, and Full Infrastructure Write Access
18h ago · 10 min read · Severity: P0 · Critical · ResolvedTechnique: Multi-turn contextual prompt chainingVulnerability Chain: Broken Access Control → Indirect Prompt Injection (RAG Poisoning) → Agentic Tool AbuseResult: Per
Join discussion
