asc-marketing-manager
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
- [PROMPT_INJECTION]: The skill processes untrusted content from Google Sheets and local asset folders to update App Store metadata. This presents an indirect prompt injection surface where malicious instructions in the source data could attempt to influence agent behavior.\n
- Ingestion points: Google Sheets (via connector) and local screenshot folder paths.\n
- Boundary markers: Data is mapped to a structural JSON schema before processing.\n
- Capability inventory: The skill can create, update, and delete metadata and screenshot sets on App Store Connect, and has network access to Apple's API.\n
- Sanitization: Character length, byte size, and content validation (e.g., Apple device name checks) are performed before data is transmitted.\n- [COMMAND_EXECUTION]: The skill instructs the agent to execute bundled Node.js scripts (scripts/asc-sync-metadata.mjs and scripts/asc-sync-assets.mjs) to interact with the App Store Connect API.\n- [DATA_EXFILTRATION]: The skill transmits app metadata and binary screenshot data to Apple's App Store Connect API. This is the intended primary function of the skill and targets a well-known and expected service.
Audit Metadata