meshy-openclaw
Pass
Audited by Gen Agent Trust Hub on Nov 10, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill installs or executes the
meshy-clipackage from the npm registry. This package is the primary tool for the skill's functionality and is maintained by the vendor. - [COMMAND_EXECUTION]: The skill executes the
meshyCLI to perform 3D generation, texturing, rigging, and print analysis. It also usesnpmfor installation and version management, and can launch local 3D slicer applications upon user request. - [DATA_EXFILTRATION]: User-provided prompts and assets (images and models) are transmitted to the Meshy API (api.meshy.ai or api.meshy.cn) for processing. This is the intended purpose of the skill for cloud-based 3D generation.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data (prompts and files) which serves as a potential attack surface. The risk is mitigated by using structured JSON communication and strictly defined execution boundaries.
- Ingestion points: User prompts, input images, and 3D models as described in
SKILL.mdandreferences/pipelines.md. - Boundary markers: The
meshyCLI returns results within a structuredmeshy.cli/v1JSON envelope, which the agent is instructed to parse. - Capability inventory: Execution of the
meshyCLI,npmpackage manager, and local slicer applications across various scripts. - Sanitization: The instructions mandate the use of structured JSON for reading command results, which prevents the interpretation of raw data as instructions.
Audit Metadata