book-sft-pipeline

Warn

Audited by Snyk on Apr 16, 2026

Risk Level: MEDIUM
Full Analysis

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

  • Third-party content exposure detected (high risk: 0.80). The skill's SKILL.md explicitly ingests ePub book content (Phase 1: extract_epub) and then reads and interprets those chunks in Phase 3 (INSTRUCTION_PROMPT with llm_call) and in dataset construction (assistant messages), so arbitrary third‑party ePubs or public datasets could contain instructions that influence segmentation, instruction generation, and downstream training actions.

Issues (1)

W011
MEDIUM

Third-party content exposure detected (indirect prompt injection risk).

Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 16, 2026, 05:24 AM
Issues
1