configure-mcp-server
Warn
Audited by Snyk on Feb 27, 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 explicitly directs adding remote MCP servers (e.g., Step 2 and Step 6: "mcp-remote https://huggingface.co/mcp" and "claude mcp add custom-server ... /path/to/server"), so the agent will connect to and consume tools/outputs from public third-party servers which could contain untrusted/user-generated instructions that influence its actions.
MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).
- Potentially malicious external URL detected (high risk: 0.90). The skill explicitly installs and runs remote code (remotes::install_github("posit-dev/mcptools") — i.e. https://github.com/posit-dev/mcptools) and configures/uses a remote MCP endpoint (mcp-remote pointing at https://huggingface.co/mcp) during setup/runtime, so fetched external content is executed/relied on and can directly control agent tool/context.
Audit Metadata