generate-image
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- Indirect Prompt Injection (LOW): The skill extracts text from local entity files in the
Worlds/directory to use as prompts for an AI image generator, which is a classic ingestion surface for indirect prompt injection.\n - Ingestion points: Markdown files located within the
Worlds/directory (specifically the## Image Promptssection).\n - Boundary markers: Absent; the skill does not use delimiters or instructions to ignore embedded commands in the extracted text.\n
- Capability inventory: File system read/write access and external image generation tool usage.\n
- Sanitization: Absent; the content following the
Prompt:marker is passed directly to the generation model.\n- No Code (SAFE): The skill consists entirely of natural language instructions (markdown) and does not contain executable scripts, shell commands, or external package dependencies.
Audit Metadata