{"repo":"komako-workshop/digital-oracle","free":true,"listed":false,"github":"https://github.com/komako-workshop/digital-oracle","clone":"git clone https://github.com/komako-workshop/digital-oracle.git","description":"AI agent skill that answers macro questions — housing, gold, BTC, geopolitics — with probability estimates mined from 13 financial data sources (Polymarket, Kalshi, CFTC, SEC & more). For Claude Code / Cursor / Codex / OpenClaw. | 让 AI Agent 从金融数据中挖掘宏观趋势的数字先知","language":"Python","stars":783,"topics":["agent-skills","ai-agent","claude-code","codex","cursor","finance","kalshi","llm","macroeconomics","market-data","openclaw","polymarket","prediction-markets","quantitative-finance"],"license":"MIT","category":"ai_agent_tool","readme_excerpt":"[English](README.en.md) | **中文**\n\n# Digital Oracle 📈\n\nDigital Oracle 是一款让 AI Agent 基于从海量金融数据中，挖掘出宏观事件发展趋势的开源 Skill。\n\n适用于 OpenClaw / Claude Code / Cursor / Codex。\n\n我们生活在一个噪音极度泛滥的时代。社交媒体上充斥着情绪化的预测 — 有人说房价要崩了，有人说黄金要上天，有人说明天就打仗。这些观点往往顺应人群情绪，而不是基于客观数据的理性分析。\n\n但交易数据不一样。\n\n价格是绝对理性的 — 当一个人要把自己的钱押在某个结果上时，他会比发一条短视频认真很多。\n\n这就是有效市场理论的核心洞察：**所有公开信息都已经被价格消化了。一切信息都在 K 线里。**\n\nDigital Oracle 把这个洞察变成了一个可执行的工具。它接入了 13 个权威金融数据源 — **从 Polymarket 和 Kalshi 这样的预测市场，到美国国债收益率曲线、CFTC 机构持仓、SEC 内部人交易、各国央行利率、加密衍生品，以及 A 股的拆单资金流与板块轮动。**\n\n它不看报纸不读新闻，不消费文章、短视频、播客，只通过从金融数据中挖掘出的价值信号，来回答房价涨跌、黄金走势、比特币周期、军事冲突概率这类问题，并给出结构化的概率估计和推理链。\n\n某种意义上，这就是新时代的数字先知。\n\n## 能回答什么问题？\n\n- \"WW3 的概率是多少？\"\n- \"中国房价还会跌多久？\"\n- \"AI 是不是泡沫？\"\n- \"现在适合买黄金吗？\"\n- \"比特币到底了吗？\"\n- \"NVDA 期权溢价是不是太高了？\"\n\n只要有市场在定价这件事，Digital Oracle 就能给出一个基于交易数据的概率估计。\n\n## 数据源\n\n| Provider | 数据类型 | 用途 |\n|----------|---------|------|\n| Polymarket | 预测市场合约 | 事件概率定价 |\n| Kalshi | SEC 监管二元合约 | 美国政治/经济事件 |\n| Stooq | 股票/ETF/外汇/商品 | 价格历史和趋势 |\n| Deribit | 加密衍生品 | 期货 term structure、期权 IV |\n| US Treasury | 国债收益率 | 利率曲线、通胀预期 |\n| CFTC COT | 期货持仓 | 机构仓位方向（smart money） |\n| CoinGecko | 加密现货 | BTC/ETH 价格、市值 |\n| SEC EDGAR | 内部人交易 | Form 4 买卖信号 |\n| BIS | 央行数据 | 政策利率、信贷/GDP 缺口 |\n| World Bank | 发展指标 | GDP、人口、贸易 |\n| Yahoo Finance | US 期权链 | IV、Greeks、put/call ratio |\n| Eastmoney | A 股行情与资金流 | 个股/ETF 行情、K 线、拆单量级资金流、板块轮动 |\n| Web Search | 网页搜索 | VIX、CDS 等补充数据 |\n\n所有 API 均免费、无需 API Key。\n\n## 安装\n\n### OpenClaw\n\n```bash\nclawhub install digital-oracle\n```\n\n### 其他 AI Agent（Claude Code / Cursor / Codex / ...）\n\n直接告诉你的 Agent：\n\n> 安装这个开源项目并读取 SKILL.md 作为你的工作指令：https://github.com/komako-workshop/digital-oracle\n\nAgent 会自行 clone 代码、阅读方法论、调用 provider。\n\n### 前置依赖\n\n- [uv](https://docs.astral.sh/uv/) — Python 包管理器，skill 运行时用它执行 Python 脚本\n- 13 个数据源中有 12 个零外部依赖（纯 Python 标准库）。期权链分析需要额外安装：\n\n```bash\nuv pip install yfinance\n```\n\n## 工作原理\n\n1. **理解问题** — 拆解核心变量、时间窗口、可定价性\n2. **选择信号** — 根据问题类型选择 3+ 个独立数据源\n3. **并行拉取** — 用 `gather()` 同时调用多个 provider\n4. **矛盾推理** — 找不同市场之间的分歧，解释为什么它们可以同时正确\n5. **输出报告** — 结构化的多层信号表格 + 概率估计 + 场景分析\n\n## 项目结构\n\n```\ndigital-oracle/\n├── SKILL.md                # Skill 定义（OpenClaw 读取这个文件）\n├── digital_oracle/         # Python 源码\n│   ├── concurrent.py       # 并行执行工具\n│   ├── http.py             # HTTP 客户端抽象\n│   ├── snapshots.py        # HTTP 响应录制/回放（测试用）\n│   └── providers/          # 13 个数据 provider\n├── references/             # API 速查\n│   ├── providers.md        # Provider API 参考\n│   └── symbols.md          # 交易符号目录\n├── scripts/                # Demo 脚本\n└── tests/                  # 单元测试 + fixtures\n```\n\n## 设计原则\n\n- **零依赖优先** — 12/13 个 provider 只用 Python 标准库，无需 `pip install`\n- **依赖注入** — 所有 provider 接受可选的 `http_client` 参数，方便测试\n- **部分失败容忍** — 一个数据源挂了不影响其他结果\n- **快照测试** — 录制真实 HTTP 响应，CI 里无网络也能跑测试\n\n## License\n\nMIT © 2026 komako-workshop — see [LICENSE](LICENSE).\n","default_branch":"main","files":76,"tree":[".gitignore","ITERATION_PLAN.md","LICENSE","README.en.md","README.md","SKILL.md","digital_oracle/__init__.py","digital_oracle/cache.py","digital_oracle/concurrent.py","digital_oracle/http.py","digital_oracle/providers/__init__.py","digital_oracle/providers/_coerce.py","digital_oracle/providers/base.py","digital_oracle/providers/bis.py","digital_oracle/providers/cftc.py","digital_oracle/providers/cme_fedwatch.py","digital_oracle/providers/coingecko.py","digital_oracle/providers/deribit.py","digital_oracle/providers/eastmoney.py","digital_oracle/providers/edgar.py","digital_oracle/providers/fear_greed.py","digital_oracle/providers/kalshi.py","digital_oracle/providers/polymarket.py","digital_oracle/providers/prices.py","digital_oracle/providers/stooq.py","digital_oracle/providers/treasury.py","digital_oracle/providers/web.py","digital_oracle/providers/worldbank.py","digital_oracle/providers/yahoo.py","digital_oracle/providers/yfinance_provider.py","digital_oracle/proxy_pool.py","digital_oracle/snapshots.py","references/providers.md","references/symbols.md","scripts/demo_deribit.py","scripts/demo_kalshi.py","scripts/demo_polymarket.py","scripts/demo_stooq.py","scripts/demo_treasury.py","scripts/regression_runner.py","tests/fixtures/deribit_book_summaries_futures.json","tests/fixtures/deribit_book_summaries_options.json","tests/fixtures/deribit_instruments_futures.json","tests/fixtures/deribit_instruments_options.json","tests/fixtures/deribit_order_book_btc_perpetual.json","tests/fixtures/kalshi_event.json","tests/fixtures/kalshi_market_detail.json","tests/fixtures/kalshi_markets.json","tests/fixtures/kalshi_orderbook.json","tests/fixtures/polymarket_book.json","tests/fixtures/polymarket_events.json","tests/fixtures/stooq_spy.csv","tests/fixtures/stooq_xauusd.csv","tests/fixtures/treasury_exchange_rates.json","tests/fixtures/treasury_nominal_curve.csv","tests/fixtures/treasury_real_curve.csv","tests/test_bis_provider.py","tests/test_cache_and_proxy.py","tests/test_cftc_provider.py","tests/test_cme_fedwatch_provider.py","tests/test_coerce.py","tests/test_coingecko_provider.py","tests/test_concurrent.py","tests/test_deribit_provider.py","tests/test_eastmoney_provider.py","tests/test_edgar_provider.py","tests/test_fear_greed_provider.py","tests/test_kalshi_provider.py","tests/test_polymarket_provider.py","tests/test_snapshots.py","tests/test_stooq_provider.py","tests/test_treasury_provider.py","tests/test_web_provider.py","tests/test_worldbank_provider.py","tests/test_yahoo_provider.py","tests/test_yfinance_provider.py"],"storefront":"/r/komako-workshop","claimed":false,"request_supported":{"post":"https://gitbuyer.com/r/komako-workshop/digital-oracle/request-supported","requests":0},"note":"indexed from public GitHub; nothing is for sale on this page. Clone it from GitHub. Paid listings live at /search."}