smart-strategy-stock-picking

Screens stocks via natural language, structured conditions, or DSL queries against QuantDB factor data.

1.5k|337|Updated Apr 15, 2026
One-click install
npx skills add https://github.com/qusong0627/QuantMind --skill smart-strategy-stock-picking-qusong0627
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: smart-strategy-stock-picking
Source: https://github.com/qusong0627/QuantMind/tree/main/skills/smart-strategy-stock-picking
Command: npx skills add https://github.com/qusong0627/QuantMind --skill smart-strategy-stock-picking-qusong0627

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Manually filtering thousands of stocks across valuation, momentum, technical, and capital-flow dimensions is slow and error-prone. This Skill turns natural language or structured conditions into executable DSL queries against the QuantDB factor store, returning a matched stock pool with quantitative metrics in seconds. ## Core Features & Use Cases - Three Query Modes: Parse free-text requests (e.g. "market cap over 50B and ROE above 15%"), structured numeric/trend/composite conditions, or raw DSL like SELECT symbol WHERE pe < 15 via the QuantMind backend API. - Rich Factor Coverage: Screens across 150+ fields from QuantDB tables including valuation (PE/PB/market cap), technical indicators (MA/RSI/KDJ/MACD), L1 factors (momentum, capital flow, concept heat), sentiment, and margin data. - Multi-Market Support: Query A-shares (SH/SZ/BJ), Hong Kong, US, and crypto markets, with summary statistics like match rate and candidate totals. - Use Case: Ask "find low-valuation blue chips with PE under 15, market cap over 50 billion, and ROE above 10%" and receive a filtered stock pool with per-stock metrics, ready for downstream news analysis, model inference scoring, or factor mining. ## Quick Start Ask the assistant to select A-share stocks where PE is below 15 and ROE is above 10 percent using the smart strategy stock picking skill.

Frequently Asked Questions about smart-strategy-stock-picking

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I screen stocks with natural language conditions?

Send your text to the /api/v1/strategy/parse-text endpoint to convert it into DSL, then execute it via /api/v1/strategy/query-pool. For example, "market cap over 50B and ROE above 15%, excluding ST stocks" becomes a SELECT symbol WHERE query.

What factors can I use for quantitative stock screening?

The skill supports 150+ fields from QuantDB tables: valuation (pe, pb, market_cap), technicals (ma5, rsi_14, macd_dif), returns (return_20d), volatility (vol_std_20), capital flow (main_flow), concept heat, industry, and ST flags.

Which markets does DSL stock selection support?

The query-pool endpoint accepts market values CN, HK, US, and CRYPTO. For A-shares you can further restrict by exchange using SH, SZ, or BJ parameters.

Why does my stock screening query return a 422 error?

A 422 means the DSL uses a field outside the backend mapping dictionary. Use the documented common factor names or check the field mappings in backend/services/engine/ai_strategy/steps/step1_stock_selection.py.

Why is my stock screening result empty?

Empty results usually mean conditions are too strict. Loosen thresholds, remove some AND clauses, and check summary.matchRate and totalCandidates to verify the candidate universe is normal.