meshyai

Pass

Audited by Gen Agent Trust Hub on Jun 13, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes shell commands (node, grep, test) to manage the Meshy AI generation workflow and verify the presence of API credentials in the local .env file.\n- [EXTERNAL_DOWNLOADS]: The skill downloads and executes the @gltf-transform/cli package via npx from the npm registry to optimize generated 3D models. As this involves a well-known service, it is documented here for visibility into the skill's network dependencies.\n- [PROMPT_INJECTION]: The skill provides an indirect prompt injection surface by interpolating user-provided text prompts and image file paths directly into shell command arguments.\n
  • Ingestion points: User input for generation prompts and image paths in SKILL.md.\n
  • Boundary markers: No protective delimiters or instructions to ignore embedded commands are present around the interpolated variables.\n
  • Capability inventory: Shell command execution via node across multiple generation and animation scripts.\n
  • Sanitization: No sanitization or input validation is performed on user-supplied strings before they are passed to the shell.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 13, 2026, 05:00 PM
Security Audit — agent-trust-hub — meshyai