The Agent Skills Directory

[SAFE]: The skill implements a comprehensive quantitative analysis framework. It fetches market data from well-known and established financial services including Tencent Finance, Sina Finance, EastMoney, and WallStreetCN. All network operations are aligned with the skill's primary purpose of providing stock quotes and market news.
[COMMAND_EXECUTION]: The skill uses a central CLI wrapper (quant.py) to execute local Python scripts for data processing, scoring, and caching. These operations are restricted to the agent's environment and a local workspace directory (~/.openclaw/workspace-trading).
[PROMPT_INJECTION]: The skill processes external news headlines for sentiment analysis. While this represents a surface for indirect prompt injection, the skill mitigates risks by stripping HTML tags and returning news as structured data to be evaluated by the agent's internal logic, rather than executing instructions contained within the news text.
[DATA_EXFILTRATION]: No evidence of unauthorized data movement. The skill primarily performs read operations from public finance APIs and write operations to its local cache for performance optimization.
[EXTERNAL_DOWNLOADS]: Dependencies listed in requirements.txt are standard and well-maintained libraries for data science and finance (pandas, pandas-ta, httpx, akshare, yfinance).

trading-quant