commerce-app-review
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONCREDENTIALS_UNSAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute the
aio auth login --barecommand to retrieve a session-based bearer token from the developer environment for authentication purposes. - [DATA_EXFILTRATION]: The retrieved Adobe IMS bearer token is transmitted to an external query endpoint at
https://commerce-docs-prod-endpoint-d0ctgyebe7bec8e6.a02.azurefd.net/api/query. Although this is used to fetch relevant documentation to enrich findings, the transmission of authentication tokens to a cloud-managed domain (Azure Front Door) is a sensitive operation. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to read and process local application files, including JavaScript logic and Markdown documentation, which represents an attack surface where malicious instructions in the analyzed code could influence the agent's review results.
- [EXTERNAL_DOWNLOADS]: The skill fetches official submission guidelines and review checklists from the
AdobeDocsorganization on GitHub, which is a trusted source for this vendor's ecosystem. - [CREDENTIALS_UNSAFE]: The provided evaluation files include example actions with hardcoded commerce API keys and secrets. These are intended as test cases for the skill to identify and are not part of the skill's own operational logic.
Audit Metadata