doubao-stock-screening

Screen and compare A-share, HK, and US stocks using structured financial data and verified business evidence.

2|Updated Aug 9, 2026
One-click install
npx skills add https://github.com/DeepJH/doubao-skill-and-info --skill doubao-stock-screening-deepjh
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: doubao-stock-screening
Source: https://github.com/DeepJH/doubao-skill-and-info/tree/main/skills/doubao-stock-screening
Command: npx skills add https://github.com/DeepJH/doubao-skill-and-info --skill doubao-stock-screening-deepjh

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? It turns open-ended stock screening questions into a traceable research workflow that defines the universe, retrieves live financial data, verifies business relevance with public disclosures, and delivers explainable candidate groupings instead of black-box rankings. ## Core Features & Use Cases - Multi-anchor screening: Handles direct condition filters, industry screening, theme/concept screening, supply-chain segment screening, strategy-style screening, stock comparison, and leader identification across A-share, HK, and US markets. - Evidence-based verification: Combines structured financial data (quotes, valuation, financials, index constituents) with public disclosures, exchange replies, and broker research to validate whether candidates genuinely benefit from a theme. - Transparent reporting: Produces formal reports with candidate tables, market/valuation comparison tables, key company deep-dives, visualizations, cited sources, and fixed risk disclaimers, delivered as Feishu documents. - Use Case: Ask "find the leading companies in the humanoid robot supply chain" and receive a scoped background analysis, a verified candidate pool grouped by evidence strength, multi-window price and valuation comparisons, and sourced conclusions without hidden scores. ## Quick Start Ask the assistant to screen stocks for a theme, industry, or strategy, for example: screen A-share semiconductor materials companies and identify the leaders with supporting evidence.

Frequently Asked Questions about doubao-stock-screening

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

FAQPage Schema
How do I screen stocks by industry, theme, or supply chain segment?

State your anchor such as an industry, concept, or supply-chain segment plus any filters like valuation or market. The skill builds a candidate pool from structured financial data, verifies business relevance through disclosures, and outputs grouped candidates with evidence and sources.

How does the skill verify a company truly belongs to a hot concept?

Platform concept tags are used only for candidate recall. Core candidates require company-level evidence such as announcements, annual reports, exchange interaction replies, orders, customers, capacity, or revenue contribution; otherwise they are moved to a watchlist.

Which markets and data sources does the stock screening support?

It covers A-share, Hong Kong, and US markets. Structured fields like quotes, valuation, financials, and index constituents come from a finance search tool, while business evidence, policies, and broker research come from general public information retrieval.

Does the skill give buy or sell recommendations or composite scores?

No. It prohibits hidden weighted scores, star ratings, theme purity percentages, and definitive advice like buy or sell. Results are transparent groupings such as core candidates, watchlist, or risk-flagged, with conditional language and disclaimers.

What happens when market data or broker research is missing?

Missing fields trigger alias matching, one round of follow-up queries, and derivation from historical prices where possible. Remaining gaps are labeled with explicit reasons like tool not returned or metric not applicable, and conclusion strength is downgraded.

What are the limitations of this stock screening workflow?

It depends on runtime retrieval tools, so coverage is limited when tools return incomplete data. Retrieval budgets cap deep-dives at three to six key companies, and outputs are informational analysis with disclaimers, not professional investment advice.