connect
Warn
Audited by Snyk on Feb 15, 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 connects the agent to many external apps (explicitly listing GitHub, Slack, Twitter, Reddit and showing examples like "Find GitHub issues labeled 'bug' from this week, summarize") and therefore the agent will fetch and read untrusted, user-generated content from public third‑party sources as part of its workflow, enabling indirect prompt injection.
Audit Metadata