cninfo-to-notebooklm

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

• Third-party content exposure detected (high risk: 0.90). The skill's required workflow and scripts (scripts/download.py and SKILL.md) explicitly fetch and download public PDFs from cninfo.com.cn (a third-party public website) and then upload those reports as sources to NotebookLM (scripts/upload.py), meaning untrusted external content is ingested and can influence the agent's analysis and downstream actions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill at runtime queries and downloads company reports from cninfo (e.g. http://www.cninfo.com.cn/new/hisAnnouncement/query and http://static.cninfo.com.cn/<adjunct_url>), then uploads those external PDFs into NotebookLM as sources—thereby injecting remote content into the model context which directly influences agent prompts/outputs and is required for the skill to function.