muapi-color-analysis-board
Pass
Audited by Gen Agent Trust Hub on May 19, 2026
Risk Level: SAFEPROMPT_INJECTIONDATA_EXFILTRATIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes an external image URL (person_image) which serves as an ingestion point for untrusted data. No sanitization, validation, or explicit boundary markers (e.g., delimiters or instructions to ignore embedded content) are present. The agent has capabilities to perform image editing (muapi image edit) and network requests (curl), creating a surface for indirect prompt injection via the processed image content.
- [DATA_EXFILTRATION]: The instructions direct the agent to perform network operations using curl to an external domain (api.muapi.ai) that is not on the standard whitelist. This includes sending an API key via headers, which constitutes external data transmission.
- [COMMAND_EXECUTION]: The skill requires the agent to execute shell commands, specifically the muapi CLI and curl, to interact with the underlying image generation service.
Audit Metadata