nano-image-generator
Pass
Audited by Gen Agent Trust Hub on Feb 18, 2026
Risk Level: SAFEPROMPT_INJECTIONCREDENTIALS_UNSAFE
Full Analysis
- Prompt Injection (LOW): The script
scripts/generate_image.pyis susceptible to indirect prompt injection as it processes unvalidated user-provided prompts that influence the model output and subsequent file system operations.\n - Ingestion points: The
promptargument is accepted via the command line inscripts/generate_image.py.\n - Boundary markers: No boundary markers or 'ignore' instructions are present in the API request construction to prevent the model from following instructions embedded within the prompt.\n
- Capability inventory: The script can perform network requests to Google APIs (
urllib.request) and write files to arbitrary local paths (Path.write_bytes), including creating directories (Path.mkdir).\n - Sanitization: No sanitization or validation is performed on the input prompt or the resulting file path beyond basic extension handling.\n- Credentials Unsafe (LOW): The script
scripts/generate_image.pypasses theGEMINI_API_KEYas a query parameter in the API URL. This practice can expose sensitive credentials in network logs, proxy servers, or browser history.
Audit Metadata