flask-review

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

• Third-party content exposure detected (medium risk: 0.30). The workflow reads human feedback and transcripts from uploaded video reviews via the Flask MCP server, allowing reviewers to inject untrusted text.

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

• Potentially malicious external URL detected (medium risk: 0.30). The skill references an MCP server endpoint at https://api.flask.do/api/mcp/mcp, which provides tool definitions and operational instructions for the agent's behavior, but lacks strong first-party verification of ownership or distribution, making it an unverifiable third-party operational dependency.