akshare-stock

Real-time and historical A-stock analysis with natural language routing across market, fundamental, sector, and derivatives data. 目标：在 OpenClaw 中通过自然语言触发 A 股和相关市场分析，输出适配 QQ/Telegram 的紧凑文本。 采用 Router -> Service -> Analyzer -> Formatter 四层结构，便于扩展和维护。 建议采用“关键词 + 正则 + 别名词典”混合方式。 下面是“功能 -> 推荐数据 -> 分析输出”的落地框架（接口以 akshare 当前版本为准，实际以 adapter 层统一封装）。 说明：以下为可直接落地的最小框架，不含完整业务细节。 建议统一为三段： 结论 -> 关键数据 -> 风险提示 。 示例： 该设计能保证你先快速可用，再逐步增强，不会一次性堆太多接口导致维护困难。 - Supports 11 analysis intents: real-time indices, K-line patterns, intraday moves, limit-up/down stats, money flow, fundamentals, margin/dragon-tiger boards, sector rotation, derivatives, funds, and cross-border markets - Four-layer architecture (Router → Service → Analyzer → Formatter) isolates intent parsing, data fetching, metric calculation, and platform-specific output formatting - Includes 30–120 second caching for market data, daily caching for financials, and fallback to last trading day data on API failures - Output automatically compresses to QQ/Telegram format with emoji, key metrics first, and 1000-character message limits with auto-splitting - 运行环境：Mac + Python 3.9 - akshare 路径： /Users/molezz/Library/Python/3.9/lib/python3.9/site-packages - Skill 入口建议： python3 skills/akshare-stock/main.py --query "${USER_QUERY}" - main.py 接收自然语言 query。 - router.py 输出结构化意图： intent + symbols + timeframe + metric + top_n 。 - services/*...