Tag: agentic-ai

All the articles with the tag "agentic-ai".

• Lakera Thingularity — Tool Recon on a Shopping Agent

  Updated: 26 Aug, 2026

  Lakera's Thingularity game hands you a full shopping agent — tools, system instructions, and business logic wired up — instead of a single LLM behind a password. Two levels in: a direct ask spills the entire tool list, then a role-manipulation and system-context-injection combo overrides a 'do not disclose' instruction after the direct approach gets refused.

• HTB Redeemer — Reading an Agent's Memory

  13 Mar, 2026

  HackTheBox Redeemer exploits unauthenticated Redis to enumerate and extract stored keys. In agentic AI systems, an open Redis instance doesn't just leak cached data — it exposes the agent's working memory, and with write access, lets an attacker inject false context the agent will act on.