create-classifier
Warn
Audited by Snyk on Mar 17, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.90). The skill explicitly runs an external "dogpile" search from scripts/assess_task.py (via DOGPILE_DIR ./run.sh search) and includes Hugging Face augmentation and scillm calls (configs/hf_augmentation.yaml and scripts/collect_bridge_data.py), so it ingests untrusted public/user-generated content (web search, HF datasets, LLM-generated variations) that the agent uses to drive model selection, training data, and routing decisions.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata