Open-Source Agentic AI Security Lab: 9 Scenarios Breaking Agent Authorization
Rewanth_Tammana · reddit · 2026-09-16
Security researcher Rewanth Tammana released 'Who Let the Agents Act?', an intentionally vulnerable agentic AI environment comparing three implementations per scenario: overprivileged, prompt-only security, and hardened with authorization enforced outside the model. Nine scenarios cover overprivileged tools, PII exposure, business-logic bypasses, indirect prompt injection, cross-tenant RAG leakage, secret propagation, fail-open authorization, agent delegation, and multi-agent confused deputy attacks. All code and execution traces are public, built on the principle that a prompt can influence an agent's behavior but should never define its authority.
More from coding & agent
- E2B Renames Its Open-Source Project to E2B Runtime, the Backend Behind Its AI Agent Cloud — badphilosopher · 2026-09-16
- Coding Agents Are Making Hypothesis Testing Cool Again in NLP Papers, Reviewer Observes — deliprao · 2026-09-16
- Dev tip: Tailscale plus Android wireless debugging is all you need — haydendevs · 2026-09-16
- ApprenticeBench: best human testers pass only 51% first-try — agents don't need perfection — hhsun1 · 2026-09-16
- God's Eye View surges on GitHub trending as community piles on new layers — bilawalsidhu · 2026-09-16
- Inspo MCP hits 1k installs; creator open-sources the Remotion launch video too — rohanpaul_ai · 2026-09-16