tushare
Pass
Audited by Gen Agent Trust Hub on Nov 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's primary purpose is to retrieve external financial data, which creates a significant surface for indirect prompt injection attacks through third-party content.
- Ingestion points: Multiple documented API endpoints (e.g., in
references/大模型语料专题数据/andreferences/大模型语料专题数据/新闻快讯(短讯).md) fetch news summaries, full-text news, company announcements, and investor-company Q&A transcripts. - Boundary markers: The implementation scripts (
scripts/stock_data_example.py,scripts/fund_data_example.py) and theSKILL.mddocumentation do not suggest using delimiters or 'ignore' instructions when processing this external text. - Capability inventory: The skill scripts demonstrate data retrieval and standard output (
print) via the Tushare client, with no evidence of dangerous write or execution operations based on the data. - Sanitization: No data sanitization or escaping mechanisms are present in the provided scripts to filter out potential injection patterns from the incoming financial data streams.
Audit Metadata