trl-fine-tuning
Warn
Audited by Snyk on Aug 22, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (medium risk: 0.30). The skill’s runtime workflow for GRPO training ingests dataset text via
templates/basic_grpo_training.py:get_dataset()(e.g.,load_dataset('openai/gsm8k', 'main')), meaning outsider-authored free-text can enter the LLM through the dataset source.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata