connect
Warn
Audited by Snyk on Feb 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: 0.80). The skill integrates with public, user-generated services (e.g., GitHub issues, Reddit, Twitter) and explicitly shows workflows like "Find GitHub issues labeled 'bug' from this week, summarize," meaning the agent fetches and reads untrusted third‑party content.
Audit Metadata