Applied AI·Security and safety
you gave the agent a write-capable tool it did not really need, and one bad step later it deleted rows you cannot get back.
Excessive agency
Draft summary, pending review
Granting a model more capability than the task warrants: write access where read would do, broad APIs, irreversible actions without confirmation. The failure amplifier behind most agent horror stories, and entry LLM06 in the OWASP list.