The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use the Bash tool to execute Python-based verification scripts located in the tools/ directory.
Evidence: python3 tools/financial_rigor.py verify-market-cap --price {价格} --shares {股本} --reported {报告市值} --currency {币种}
Risk: The instructions direct the agent to populate shell commands with variables (e.g., {价格}, {股本}) derived from research conducted by subagents on the web. The absence of instructions for sanitizing or escaping these values creates a command injection surface if the external financial data contains malicious shell metacharacters.
[EXTERNAL_DOWNLOADS]: The skill instructions direct subagents to retrieve financial data from several external websites.
Evidence: The prompt templates for subagents specify using WebSearch to query macrotrends.net, stockanalysis.com, aastocks.com, eastmoney.com, cninfo.com.cn, and goodinfo2.tw.
Note: These sources are well-known and established financial data providers.
[DATA_EXPOSURE]: The skill workflow includes writing final reports to the user's home directory.
Evidence: ~/{公司名}投资研究报告_{日期}.md
Note: While writing to the home directory is a sensitive operation, it is contextually aligned with the primary purpose of the skill to generate and save investment reports for the user.

investment-team