character-design-sheet
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references an external CLI tool ('belt') and provides a link to installation instructions hosted on GitHub ('https://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md'). These are documented as requirements for the image generation functionality.
- [COMMAND_EXECUTION]: The skill's instructions involve using the 'belt' CLI tool to run various AI models for image generation and processing (e.g., 'falai/flux-dev-lora', 'infsh/stitch-images'). This tool is explicitly permitted in the skill's 'allowed-tools' metadata.
- [INDIRECT_PROMPT_INJECTION]: The skill utilizes templates that interpolate user-supplied character descriptions directly into prompts for AI image generation models.
- Ingestion points: User-provided character descriptions used in the 'Description Anchor' templates (SKILL.md).
- Boundary markers: No specific delimiters or safety warnings are included in the prompt templates to prevent the agent from following instructions embedded within user character descriptions.
- Capability inventory: The skill utilizes 'belt app run' which performs network operations to interact with external AI inference APIs.
- Sanitization: There is no evidence of sanitization or validation of the user-provided text before it is interpolated into the final prompt command.
Audit Metadata