Daily Stock Analysis Tool Features: AI-Driven Stock Research Platform Explained

The daily stock analysis tool is a full-stack, AI-driven platform that automates stock research across global markets (A-share, Hong Kong, US), generates actionable investment dashboards, and delivers multi-channel notifications through a modular, extensible architecture.

The ZhuLinsen/daily_stock_analysis repository provides a comprehensive solution for quantitative and fundamental analysis. This open-source tool combines large language models with real-time market data to produce daily reports, back-test investment strategies, and enable conversational AI agents for stock research.

AI-Powered Decision Dashboard

At the core of the system is the AI Decision Dashboard, which generates a one-sentence investment conclusion, numerical score, buy/sell points, risk alerts, and an actionable checklist for every analyzed security.

The dashboard construction logic resides in src/notification.py, where the NotificationService class assembles markdown reports by aggregating technical indicators, sentiment analysis, and fundamental data into a unified AnalysisResult structure. This service supports both individual stock analysis and batch processing workflows.

Multi-Market Analysis Capabilities

The tool provides global market coverage supporting A-shares, Hong Kong stocks, US equities, US indices, and common ETFs. Market region selection is handled dynamically through the MARKET_REVIEW_REGION configuration in src/core/market_review.py.

The multidimensional analysis engine orchestrates data collection through src/analyzer.py, which coordinates:

• Technical indicator calculation
• Real-time quote fetching
• Chip-distribution analysis
• News sentiment extraction
• Capital flow tracking
• Fundamental data integration

All data streams merge into a unified analysis result before LLM enhancement.

Strategy Engine and AI Agent

A YAML-driven strategy system allows custom algorithm implementation alongside 11 built-in strategies including A-share recap, US Regime detection, moving-average golden-cross, Chan-Lun theory, wave theory, and sentiment cycle analysis. Strategy definitions follow the format documented in strategies/README.md.

The Agent/Strategy Conversation feature enables multi-turn dialogue that invokes specific strategies via natural language. The entry point in src/agent/strategies/strategy_agent.py handles requests such as analyzing specific stocks through particular technical lenses, making the tool accessible to non-technical users.

Data Architecture and Real-Time Features

The extensible data provider layer (data_provider/base.py) implements a unified DataFetcherManager interface that abstracts multiple data sources including TickFlow, AkShare, Tushare, YFinance, and Longbridge. This architecture supports automatic fallback and priority rules when primary sources are unavailable.

For US equities, the optional Social Sentiment Service (src/services/social_sentiment_service.py) aggregates Reddit, X (Twitter), and Polymarket data via external APIs. The pipeline also supports real-time quote and chip distribution fetching, configurable through enable_realtime_quote and enable_chip_distribution flags in src/core/pipeline.py.

Automation and Scheduling

The tool supports multiple deployment modes through src/scheduler.py:

• GitHub Actions integration for cloud execution
• Docker containerization for consistent environments
• Local cron-style scheduling for on-premise deployment
• Permanent FastAPI service mode for API access

Configuration is environment-driven via src/config.py, which supports runtime reload for scheduled operations without process restart.

Web Interface and Report Management

The dual-theme workbench (src/webui_frontend.py) provides manual analysis capabilities, configuration management, task progress monitoring, historic report browsing, and portfolio management through a FastAPI-based interface.

Intelligent import features in src/services/name_to_code_resolver.py accept images, CSV/Excel files, and clipboard data while auto-completing stock codes, names, pinyin, or aliases. Every generated report persists to a SQLite database via src/services/history_service.py, enabling re-analysis and batch export functionality.

AI Back-Testing and Verification

The AI-Backtest Verification system (src/services/backtest_service.py) re-runs historic analyses to compare AI decisions against actual market outcomes. This capability computes simulated returns and validates strategy effectiveness across different market conditions.

Users can trigger back-tests via CLI:

python main.py --backtest --backtest-code 600519 --backtest-days 30

Multi-Channel Notification System

Completed analyses distribute through multi-channel notifications implemented in src/notification_sender/. Concrete implementations include:

• wechat_sender.py for WeChat Work integration
• telegram_sender.py for Telegram bots
• Feishu, Discord, Slack, Email, Pushover, and custom webhook support

The system pushes either full dashboards or individual stock notes based on configuration.

Command-Line Interface Examples

Run single-stock analysis

python main.py --stocks 600519,hk00700,AAPL

Generate market review only

python main.py --market-review

Start FastAPI server

python main.py --serve-only

Enable scheduled daily runs

export SCHEDULE_ENABLED=true
export SCHEDULE_TIME=18:00
python main.py --schedule

Use strategy agent via API

curl -X POST http://127.0.0.1:8000/api/v1/agent/chat \
     -H "Content-Type: application/json" \
     -d '{"message":"Analyze 600519 using moving average strategy"}'

Summary

• ZhuLinsen/daily_stock_analysis provides AI-enhanced stock analysis across A-share, Hong Kong, and US markets through a modular Python architecture.
• The platform features 11 built-in trading strategies with YAML-based customization, real-time data fetching, and social sentiment integration for US stocks.
• Users access functionality via CLI automation, scheduled tasks, FastAPI endpoints, or a dual-theme web interface with intelligent data import capabilities.
• Comprehensive back-testing in src/services/backtest_service.py validates AI decisions against historical market performance.
• Multi-channel notifications support enterprise messaging platforms including WeChat Work, Feishu, Telegram, and Slack.

Frequently Asked Questions

What markets does the daily stock analysis tool support?

The tool supports A-shares (China mainland), Hong Kong stocks, US equities, US indices, and common ETFs across all major exchanges. Market selection is configured through the MARKET_REVIEW_REGION environment variable or flag, processed in src/core/market_review.py.

How does the AI strategy agent work?

The strategy agent in src/agent/strategies/strategy_agent.py enables natural language interaction with 11 built-in technical analysis strategies. Users can request specific analyses like "check for golden-cross on 600519" via the FastAPI endpoint or CLI, and the system translates these into structured strategy invocations with LLM-enhanced interpretations.

Can I run this tool without programming knowledge?

Yes. While the tool requires Python installation, the FastAPI web interface (src/webui_frontend.py) provides a graphical workbench for manual analysis, configuration management, and report browsing. Additionally, the intelligent import system in src/services/name_to_code_resolver.py handles fuzzy stock name matching, supporting pinyin and aliases for easier lookup.

What data sources does the tool use?

The DataFetcherManager in data_provider/base.py abstracts multiple providers including AkShare, Tushare, YFinance, TickFlow, and Longbridge, with automatic fallback logic. For US stocks, optional social sentiment data from Reddit, X, and Polymarket integrates through src/services/social_sentiment_service.py.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}