小龙虾 A股热点助手

Purpose

Use this skill for the project’s “小龙虾” stock-hotspot workflow.

Workflow

1. Load environment variables from .env.
2. Collect market data with fetch_market_hotspots() in agent.py.
3. Build the analysis prompt with build_prompt().
4. Call analyze_with_glm() with ZHIPUAI_API_KEY.
5. Present results in two parts:
   • 原始热点数据 JSON
   • GLM-4 板块方向建议

Output style

• 语气要克制、谨慎、研究化。
• 输出要结构清晰，优先使用分点和小标题。
• 始终保留“仅供研究，不构成投资建议”的免责声明。
• 如果用户要求“更像小龙虾”的表达，可以在不改变专业性的前提下，稍微更亲切、轻松一点。

Key implementation notes

• Use AkShare for index and concept-board snapshots.
• Sort concept boards by 涨跌幅 and 成交额.
• Handle data-fetch failures gracefully and include error fields in the returned JSON.
• Keep prompts aligned with the existing system message: structured, cautious, and compliant.
• If the user asks for prompt tuning, preserve the same research constraints and disclaimer.

Common edits

When changing this project, usually update one of these:

• agent.py for data collection or prompting logic
• README.md for usage instructions
• requirements.txt for dependencies

Example use

• 运行项目并总结今天的小龙虾热点板块观察。
• 调整 top_n 改变分析的板块数量。
• 在发送给 GLM-4 之前增加更多市场数据源。