MCP Architecture Expert

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly describes and examples MCP servers that expose and read from public, user-generated sources (e.g., "Common MCP Servers" listing GitHub, Slack, Google Drive, and "Puppeteer
• Web scraping", plus resource examples like "github://issues" and file/db URIs) which the agent is expected to read/interpret and could materially influence tool invocation and actions, enabling indirect prompt injection.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill includes runtime commands that fetch and execute remote packages (e.g., "npx @modelcontextprotocol/server-github", "pip install mcp-server-slack", and "docker run mcp-postgres-server"), which will run external code that can provide MCP prompts/tools to the agent and therefore directly influence agent behavior.