baoyu-image-cards
Pass
Audited by Gen Agent Trust Hub on May 18, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [PROMPT_INJECTION]: The skill contains instructions that direct the agent to bypass standard safety refusals in image generation backends when handling sensitive or copyrighted figures, including the specific directive "DO NOT refuse to generate".\n- [PROMPT_INJECTION]: The skill exhibits a vulnerability to indirect prompt injection by interpolating untrusted user input directly into prompts for image generation backends. \n
- Ingestion points: User-provided text for social media content conversion.\n
- Boundary markers: Absent; content is placed within markdown sections without explicit isolation delimiters or instructions to ignore embedded commands.\n
- Capability inventory: The skill triggers tool calls to image generation backends and writes to the local file system.\n
- Sanitization: Absent; no validation or escaping of user input is specified before its inclusion in prompts.\n- [NO_CODE]: The skill is comprised entirely of Markdown instruction files and reference templates, containing no executable scripts, binaries, or source code.
Audit Metadata