douyin-cover
Pass
Audited by Gen Agent Trust Hub on Aug 13, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: Indirect Prompt Injection Surface. The skill ingests user-provided video titles and existing covers for generation and diagnosis, which could allow malicious inputs to influence the agent's execution flow.\n
- Ingestion points: User-provided titles and cover images processed in SKILL.md and references/cover-diagnosis.md.\n
- Boundary markers: No explicit delimiters or instructions are used to isolate user input from the skill's logic.\n
- Capability inventory: The skill uses image generation tools, image viewing tools, and shell commands (mkdir, cp) for file organization.\n
- Sanitization: User input is not explicitly filtered or sanitized before being used in the image generation process.\n- [COMMAND_EXECUTION]: Local File Management. The skill uses standard shell commands like
mkdirandcpto manage assets within the local filesystem, specifically for organizing generated images into project directories as defined in SKILL.md.
Audit Metadata