create-mcp-skill

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 using MCP servers (e.g., chrome-devtools-mcp) and mcp shell to run commands like navigate_page {"url":"https://example.com"} and to "discover" and "test" tools—showing the agent will fetch and interpret public web content (example.com / web navigation) as part of its workflow, which could let untrusted page content influence subsequent tool calls.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill includes runtime commands that fetch and execute remote packages (e.g., "go install github.com/f/mcptools/cmd/mcptools@latest" and runtime package invocations like "bunx ...@latest" / "npx @modelcontextprotocol/server-filesystem") — for example the GitHub source https://github.com/f/mcptools is pulled/installed and used at runtime, which means external code is fetched and executed as a required dependency.