idea-generation

SKILL.md 节选

# idea-generation

## Purpose

Systematically surface new **全球股票市场（美股/港股/A 股为主，兼顾其他市场）投资机会** through quantitative screens, thematic analysis, and pattern recognition.

## Data Sources

Two MCP connectors cover the full sourcing workflow:

- `valuz-stock` (Stock MCP) — 行情、财务、指标、因子筛选数值数据. 用**裸代码** (`AAPL` / `00700` / `600519`).
- `valuz-search` (Search MCP) — 财报、公告、研报、纪要、电话会检索. 用 **`market:ticker`** 格式 (`US:AAPL` / `HK:00700` / `SH:600519`).

> **Symbol formats** — `valuz-stock` 用裸代码；`valuz-search` 用 `market:ticker`；`factors_screen` 用 `market` 参数 (`cn` / `hk` / `us`).

用 `valuz-stock` 的因子引擎 (`factors` / `factors_compute` / `factors_screen`) 做量化筛选，用 `valuz-search` 取定性催化/主题资料。

### Primary: valuz-stock

```python
factors()                                     → 列出可用因子 (PE/PB/ROE/RSI/MACD…)
factors_compute(symbol, formula)              → 算单只标的因子值, 验证公式
factors_screen(market, formula)               → 因子选股 (核心), market=cn/hk/us
stock_quote(symbol)                           → Price, PE, PB, market cap
index_quote(symbol)                           → 指数行情
company_overview(symbol)                       → 公司概览
income_statement(symbol, period, limit)       → Financial metrics
balance_sheet(symbol, period, limit)          → Balance sheet health
cashflow_statement(symbol, period, limit)     → 现金流
revenue_breakdown(symbol)                      → 收入拆分
industry_constituents(...)                     → Peer comparison
concepts_today() / concepts_latest()           → 当日/最新热门概念
ohlcv(symbol) / kline(symbol)                  → Price trends, momentum
backtest(...)                                  → 策略回测
```

### Secondary Screening Data

| Data | Source | Use |
|------|--------|-----|
| Hot concepts / themes | valuz-stock (`concepts_today`, `concepts_latest`) | Theme rotation, where the money is going |
| Index / industry constituents | valuz-stock (`index_constituents`, `industry_constituents`) | Universe definition, peer set |
| Shareholder structure | valuz-stock (`company_shareholders`) | Institutional accumulation/distribution |
| Earnings calendar | valuz-stock (`earnings_calendar`) | Catalyst timing |
| Institutional / analyst views | valuz-search (`reports_search`) | Sell-side conviction, target prices |
| Filings & disclosures | valuz-search (`filings_search`) | Insider transactions, lock-up, M&A |

## Workflow

### Step 1: Define Screen Criteria

**Investment philosophy alignment:**
- Value vs Growth vs GARP vs Momentum
- Market cap preference (large / mid / small)
- Sector focus or sector-agnostic
- Liquidity requirements (turnover threshold)
- Risk tolerance (volatility, leverage, earnings stability)

**Screen parameters (market-aware, fields not tied to one market's data vendor):**

| Parameter | Typical Range | Notes |
|-----------|--------------|-------|
| PE (TTM) | 5-50x | Avoid negative PE |
| PB | 0.5-5x | <1x may indicate distress |
| PS | 0.5-5x | For high-growth unprofitable |
| Market cap | above liquidity threshold | Tradability |
| Daily turnover | above liquidity threshold | Tradability |
| ROE | >10% | Quality filter |
| Debt/Equity | <100% | Financial health |
| Revenue growth | >10% | Growth filter |
| EPS growth | >15% | Earnings momentum |

### Step 2: Quantitative Screens

筛选用 `valuz-stock` 的 **`factors_screen`** (核心)，传 `market` (`cn` / `hk` / `us`) + 因子 `formula`。
先用 `factors()` 看可用因子，必要时用 `factors_compute(symbol, formula)` 在单只标的上验证公式再放大到全市场。

> **Factor syntax** — 技术: `RSI(14)` / `MACD()` / `BOLL(20,2)` / `KDJ()` / `ATR(14)`；
> 基本面: `PE()` / `PE_TTM()` / `PB()` / `ROE()` / `ROA()` / `EPS()`；
> 财报字段: `INCOME.net_profit` / `BALANCESHEET.total_assets`；

SKILL.md 不止这些内容，安装后可阅读全文。

技能包内容

共 1 个文件，合计 13.3 KiB