takedown
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection because it is designed to ingest and analyze external, untrusted documents (legal notices) provided by users or third parties.
- Ingestion points: In the "接收模式" (Receive Mode), the skill reads and triages external files, such as the example
~/Downloads/平台通知.pdf. - Boundary markers: The instructions do not specify any delimiters or safety markers to isolate the content of the untrusted notice from the agent's internal instructions.
- Capability inventory: The skill has the capability to write files (memos and drafts) to the local workspace ("写入事项文件夹").
- Sanitization: There are no instructions provided to sanitize, escape, or filter the content extracted from the external legal notices before processing it.
Audit Metadata