image-element-remove

Pass

Audited by Gen Agent Trust Hub on Aug 29, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes the meitu command-line interface. This is a vendor-provided tool used to interface with the Meitu OpenAPI for image processing. The tool is invoked to perform element removal based on user-provided parameters.\n- [EXTERNAL_DOWNLOADS]: The instructions guide the user to install or update the meitu-cli package via the NPM registry. This is the official tool for the service provided by the vendor. The CLI interacts with Meitu's official OpenAPI endpoints for image processing.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied image descriptions, creating an ingestion point for untrusted data in SKILL.md. It uses a structured four-segment English prompt template (incorporating SCOPE, PRESERVE, REFLOW, and NEGATIVE sections) as a boundary to guide the image editing model. The available capabilities are restricted to executing the meitu CLI and writing processed images to specified local directories (./output/ and workspace paths). No explicit sanitization of the user description is mentioned beyond the template structure.\n- [DATA_EXPOSURE]: The skill accesses API credentials stored in ~/.meitu/credentials.json or provided via environment variables such as MEITU_OPENAPI_ACCESS_KEY. This access is explicitly declared in the skill's metadata and requirements for the purpose of authenticating with the vendor's image processing service.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 29, 2026, 02:53 AM
Security Audit — agent-trust-hub — image-element-remove