meme-generator
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to run a shell command invoking a local Python script (
skills/shared/scripts/meme_ops.py). It uses user-provided input for the output path (outputs/主题名/), which could potentially be used for path traversal if the agent does not sanitize the "theme name" provided by the user. - [INDIRECT_PROMPT_INJECTION]: The skill takes untrusted user input and interpolates it directly into shell command arguments, creating a potential command injection surface.
- Ingestion points: User-supplied strings for meme labels (
--top,--bottom,--caption) and the theme name for the output directory. - Boundary markers: Absent. No instructions are provided for the agent to validate, delimit, or escape characters in the user input.
- Capability inventory: Shell command execution via
python skills/shared/scripts/meme_ops.py. - Sanitization: Absent. The skill instructions do not provide logic for escaping shell metacharacters (e.g.,
;,&,|,`) or directory traversal sequences (e.g.,../).
Audit Metadata