yao-geo-xiaohongshu
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill implements proactive security measures in
scripts/build_workbook.py, including an escaping mechanism for spreadsheet data to prevent formula injection and directory traversal checks for local file paths. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external content from search suggestions, but includes mitigations to handle this safely.
- Ingestion points: Xiaohongshu autocomplete suggestions are retrieved via browser tools.
- Boundary markers: The
SKILL.mdinstructions explicitly direct the agent to treat webpage text as inert data and ignore any instructions found within it. - Capability inventory: The skill performs local file system writes for data persistence and Excel generation.
- Sanitization: The
safe_cellfunction ensures that all external text is sanitized before being written to the final Excel report.
Audit Metadata