moonkite-wechat-hongbao
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION] (LOW): The
promptargument is passed directly to the Gemini API. This is the intended functionality but represents a standard surface for prompt injection if the agent processes untrusted user input. - [DATA_EXPOSURE] (LOW): The script facilitates reading local files (via
--reference) and writing to the local filesystem (via--output). While appropriate for a local utility, if deployed in a multi-user or automated environment, path validation should be implemented to prevent unauthorized file access. - [EXTERNAL_DOWNLOADS] (SAFE): The script depends on the legitimate
google-genailibrary to communicate with Google services. No unauthorized third-party downloads or remote script executions were found.
Audit Metadata