ip-character-sheet
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill relies on local Python scripts (
generate.py,make_board.py,palette.py) to perform its primary functions, including image generation requests, layout assembly for brand assets, and color palette extraction. - [EXTERNAL_DOWNLOADS]: The
generate.pyutility connects to external API endpoints (such as OpenAI or user-specified gateways like tryallapi.com) to transmit design prompts and retrieve generated image files. - [INDIRECT_PROMPT_INJECTION]: The skill constructs image generation prompts by interpolating user-supplied brand information and personality keywords. While there is a surface for indirect injection, the impact is confined to the quality and content of the generated image output.
- Ingestion points: User-provided brand names, positioning statements, and keywords processed via templates in
SKILL.md. - Boundary markers: None identified in the prompt templates.
- Capability inventory:
scripts/generate.py(Network access, local file writing),scripts/make_board.py(Local file reading and writing). - Sanitization: No specific filtering or escaping logic is applied to user-provided text before prompt interpolation.
Audit Metadata