gpc-review-management
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-generated reviews from the Google Play Store, which represents an ingestion of untrusted external data.
- Ingestion points: Data is ingested into the agent context through the outputs of the
gpc reviews listandgpc reviews getcommands defined inSKILL.md. - Boundary markers: The instructions do not specify the use of delimiters or boundary markers to separate user-provided review text from the agent's instructions.
- Capability inventory: The skill possesses a network-based write capability via the
gpc reviews replycommand, which allows the agent to post content to the Google Play Store. - Sanitization: There is no evidence of sanitization or filtering of the review content before it is processed by the agent or displayed to the user.
- Mitigation: The skill explicitly instructs the agent to draft replies for user approval and prohibits automatic replies, providing a human-in-the-loop checkpoint that mitigates the risk of unintended actions triggered by malicious review content.
Audit Metadata