gift-up
Pass
Audited by Gen Agent Trust Hub on Mar 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
membraneCLI to perform administrative tasks, manage connections, and execute actions against the Gift Up! API. - [EXTERNAL_DOWNLOADS]: Recommends the installation of the
@membranehq/clipackage, a vendor-provided tool required for the skill's functionality. - [PROMPT_INJECTION]: The skill processes external data retrieved from the Gift Up! API, creating a surface for indirect prompt injection where malicious instructions in the data could influence agent behavior.
- Ingestion points: API responses from actions such as
list-items,get-gift-card, andlist-gift-cardsdefined inSKILL.md. - Boundary markers: No explicit boundary markers or instructions to ignore embedded commands are specified for the ingested data.
- Capability inventory: The skill possesses state-changing capabilities, including
void-gift-card,redeem-gift-card, and arbitrarymembrane requestexecution. - Sanitization: No sanitization or validation logic is defined for the content returned by the Gift Up! service.
Audit Metadata