The Agent Skills Directory

[SAFE]: The skill implements a strong security posture for financial operations. It explicitly mandates a human-in-the-loop workflow in references/trade.md, requiring the agent to present order details in a table and wait for explicit confirmation before executing place_order(). It also defaults to Paper Trading to prevent accidental real-money loss.
[SAFE]: Sensitive credential management follows best practices. The skill instructs users to store API keys and RSA private keys in a local configuration file (tiger_openapi_config.properties) rather than hardcoding them within the code or prompts.
[EXTERNAL_DOWNLOADS]: The skill identifies necessary build dependencies (Boost, Protobuf, Abseil, cpprestsdk) and provides instructions to clone the official SDK from the author's GitHub organization (github.com/tigerfintech). These are legitimate development resources associated with the vendor.
[INDIRECT_PROMPT_INJECTION]: While the skill provides capabilities to ingest market data and execute trades based on user input, the risk of indirect injection is mitigated by the mandatory confirmation steps and the use of structured order objects via OrderUtil and ContractUtil, which prevents the mixing of instructions and data in API calls.

tigeropen-cpp