fine-tuning-openvla-oft
Pass
Audited by Gen Agent Trust Hub on Mar 19, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [SAFE]: The skill provides a detailed reproduction guide for the OpenVLA-OFT research paper. The instructions and code snippets follow standard machine learning development practices for environment configuration and model execution.\n- [EXTERNAL_DOWNLOADS]: The skill automates the retrieval of source code from public GitHub repositories (moojink/openvla-oft and Lifelong-Robot-Learning/LIBERO) and model weights or datasets from Hugging Face. These are recognized and established sources within the robotics research community.\n- [COMMAND_EXECUTION]: Provides instructions to execute shell commands for package management (conda, pip) and training orchestration (torchrun). These actions are necessary for the primary purpose of the skill.
Audit Metadata