cn-finance-data
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted text data from external financial news feeds and investor Q&A platforms (e.g., irm_qa_sh, irm_qa_sz). This represents a surface for indirect prompt injection attacks where malicious instructions could be embedded in the fetched content.
- Ingestion points: External text data is retrieved via API methods in the tushare library and handled by scripts/api_client.py.
- Boundary markers: No specific boundary markers or instructions were found to separate untrusted text data from the agent's internal logic.
- Capability inventory: The skill has access to the Bash tool for running Python scripts, which increases the impact of potential injections.
- Sanitization: The skill does not perform sanitization on the external text data before it is presented to the agent.
- [EXTERNAL_DOWNLOADS]: The skill connects to the well-known tushare.pro service to fetch financial market data, which is consistent with its primary purpose and functionality.
Audit Metadata