paste-up

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

• Third-party content exposure detected (medium risk: 0.30). The workflow reads design specifications, project docs, and Paper file comment threads via MCP tools, which can contain outsider-authored text.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill configures an operational MCP dependency pointing to a local URL (http://127.0.0.1:29979/mcp), which falls under weak/unknown provenance due to the raw loopback IP address.