agent-reach
Warn
Audited by Snyk on Sep 10, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (medium risk: 0.30). The required workflow instructs the agent to run third-party platform CLI tools and MCP servers (such as Twitter, GitHub, Reddit, LinkedIn, and Exa search) that return untrusted outsider-authored posts, comments, and search results into the agent runtime context.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata