findata-toolkit-cn
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external financial data sources via the AKShare library. While external data ingestion is a potential surface for indirect prompt injection, the skill specifically targets structured market data (OHLCV, financial ratios, macro indicators) and includes safe parsing utilities to handle missing or malformed values.
- Ingestion points: External data is fetched through AKShare in
scripts/stock_data.pyandscripts/macro_data.py. - Boundary markers: The skill outputs structured JSON to standard output, providing a clear boundary for the agent.
- Capability inventory: The skill performs network operations to financial websites (via AKShare) and reads a local configuration file.
- Sanitization: The implementation includes
_normalize_symbolfor stock codes andsafe_floatfor data parsing to prevent processing of malformed inputs.
Audit Metadata