summarization

Warn

Audited by Snyk on Feb 28, 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.md explicitly instructs the agent to fetch arbitrary web content with "WebFetch if a URL is provided" (Step 1: Identify the Source), and that fetched, untrusted web content is read/parsed and used to generate summaries, action items, and recommendations—allowing third-party text to influence the agent's outputs and next actions.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 28, 2026, 08:16 AM