comms-ops
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements strict data integrity controls, mandating the use of
[NEEDS INPUT]markers instead of allowing the agent to invent metrics, customer names, or business outcomes. - [SAFE]: A strong human-in-the-loop constraint is established; the instructions explicitly state "Draft only—never publish, never send, never speak," ensuring the AI agent cannot autonomously communicate with third parties.
- [SAFE]: The strategist instructions include specific compliance rules for media platforms (such as Qwoted and Featured), prohibiting the submission of unreviewed AI-generated quotes to prevent terms-of-service violations and reputation damage.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external news stories and internal brand context files. While this constitutes an attack surface for indirect prompt injection, the risk is mitigated by the absence of high-privilege tool access (e.g., shell execution) and the mandatory human review of all deliverables.
Audit Metadata