project-setup-info-local

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

• Third-party content exposure detected (high risk: 1.00). The skill's SKILL.md explicitly requires visiting the public GitHub org (https://github.com/modelcontextprotocol) and using the fetch_webpage tool to retrieve and act on content from the public URL https://modelcontextprotocol.io/llms-full.txt, meaning the agent will ingest and use untrusted third-party webpage content to drive setup decisions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly requires visiting https://github.com/modelcontextprotocol and using the fetch_webpage tool to retrieve https://modelcontextprotocol.io/llms-full.txt at runtime to obtain implementation instructions that are then incorporated into the project's copilot-instructions (i.e., remote content directly controls prompts/instructions and is a required dependency).