pdf-ocr-skill

Warn

Audited by Snyk on Mar 15, 2026

Risk Level: MEDIUM
Full Analysis

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

  • Potentially malicious external URL detected (high risk: 0.70). The skill's SiliconFlow engine makes runtime POST requests to https://api.siliconflow.cn/v1/chat/completions and consumes the model's responses as the OCR output (i.e., external model-generated text directly controls the skill's returned content), which is a runtime external dependency that can influence agent behavior.

Issues (1)

W012
MEDIUM

Unverifiable external dependency detected (runtime URL that controls agent).

Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 15, 2026, 08:51 AM
Issues
1