pytorch-lightning
Warn
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- EXTERNAL_DOWNLOADS (MEDIUM): The skill instructs users to install the 'lightning' package. Neither the author 'Orchestra Research' nor the 'Lightning-AI' organization are present in the defined list of trusted organizations.
- PROMPT_INJECTION (LOW): The skill identifies a surface for indirect prompt injection. 1. Ingestion points: Untrusted data enters the agent context through the 'train_loader' and 'val_loader' objects. 2. Boundary markers: No specific delimiters or safety instructions are used to distinguish training data from system instructions. 3. Capability inventory: The skill performs file system operations including model checkpointing and logging. 4. Sanitization: No input validation or sanitization of the training data is implemented.
- COMMAND_EXECUTION (SAFE): Shell commands are restricted to standard package installation and python script execution for the primary training task.
Audit Metadata