The Agent Skills Directory

[SAFE]: No malicious code or security vulnerabilities were detected in the skill's instruction files. The skill manages data retrieval by running local scripts (e.g., ../findata-toolkit-cn/scripts/views_runner.py) using a defined virtual environment, which is a secure method for executing internal tools. There are no hardcoded credentials; the reference to XUEQIU_TOKEN is a configuration instruction for the user. While the skill processes external financial data, it utilizes structured analysis templates that define safe boundaries. The use of relative paths for script execution and requirements installation is legitimate within the context of the provided toolkit. The capability inventory is restricted to intended research functions, and no evidence of data exfiltration or prompt injection was found.

event-study