tech-plan-assessment

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly runs a runtime download of user-provided Feishu documents via the feishu2md command (e.g. URLs matching https://.feishu.cn/docx/ or https://.feishu.cn/wiki/) and then ingests that fetched Markdown as the PRD used to drive the agent's evaluation, so external content would directly control the agent's prompts/context.