Topic
Threats to autonomous agents that read external content, including indirect prompt injection.
No current published clusters are mapped here yet.