implementing-llms-litgpt
Warn
Audited by Snyk on Jun 22, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.70). Outsider free text can enter the LLM context via the user-provided fine-tuning dataset JSON (
--data.json_path), whoseinstruction/input/outputfields are arbitrary text authored by whoever created that dataset, and LitGPT will tokenize and feed it to the model during training.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata