generate-characters
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection by processing external data and using it to form AI prompts.\n
- Ingestion points: Character descriptions and style definitions are read from
project.jsonand script files inprojects/{project_name}/scripts/(referenced inSKILL.mdandgenerate_character.py).\n - Boundary markers: No explicit boundary markers or instructions to ignore embedded commands were detected in the prompt construction logic.\n
- Capability inventory: The skill can perform file system writes (saving images), modify local project configuration files, and make network requests to image generation APIs.\n
- Sanitization: The skill lacks mechanisms to sanitize or validate the character descriptions before passing them to the Gemini API client.
Audit Metadata