Daily A-Share Analysis

Background

Tracking daily A-share market movements and identifying trading opportunities is a time-consuming and tedious task. Investors need to constantly analyze market data, screen potential stocks, and formulate detailed trading execution plans. Manual handling of these tasks is not only inefficient but also prone to human errors that may lead to missed optimal timings. This skill aims to automate the process, providing structured analysis support.

Core Working Principle

The skill addresses these issues through an automated workflow. The core process includes:

• Data Acquisition and Caching: The skill prioritizes using cached rating data from the local file a_share_scores.csv to avoid unnecessary network requests. Unless the user explicitly requests a market data refresh, it uses the cache to ensure quick response.
• Candidate Stock Screening: Based on predefined stock selection logic, the skill screens 5 trading candidate stocks worth tracking from the entire market data. Screening criteria cover technical indicators and fundamental factors, with specific parameters adjustable via workspace scripts.
• Trading Plan Generation: For each candidate stock, the skill calculates breakout entry points, pullback entry points, stop-loss levels, and take-profit levels. These price levels are generated based on real-time market data or fallback template modes and output to the execution plan file a_share_agent_execution_plan.csv.
• Report Output: The skill generates a daily analysis report in Markdown format stored in reports/a_share_daily_report_YYYYMMDD.md. The report includes market overview, candidate stock analysis, and trading recommendations, using **bold** to emphasize key data.
• Condition Order Export: To facilitate manual entry into brokerage systems, the skill exports a condition order sheet to a_share_manual_order_sheet.csv, allowing traders to operate directly and reduce manual entry errors.
• Scheduled Task Support: The skill can be set to execute regularly on both Windows and Linux environments. On Windows, it registers via PowerShell scheduled task scripts; on Linux, it prioritizes OpenClaw cron or system cron for daily automated report generation.

Considerations

The following boundaries should be noted when using the skill:

• Data Dependencies: The skill relies on market data sources. If real-time quotes or historical data fetching fails, it falls back to template mode, but this may affect analysis accuracy. Users should ensure data interfaces are functioning properly or handle network exceptions.
• Environment Compatibility: The skill has different execution paths and scripts on Windows and Linux, such as .venv/Scripts/python.exe on Windows and .venv/bin/python on Linux. Users need to select the correct commands based on the operating system to avoid path errors.
• Push Capabilities: Currently, the skill only generates local files, such as Markdown reports and CSV files, and does not integrate email, webhook, or other external message push mechanisms. For real-time notifications, users need to configure this themselves or use additional tools.
• Configuration Customization: Unless the user explicitly requests changes to thresholds, factors, or screening criteria, the skill uses the existing stock selection logic. Custom adjustments require editing workspace scripts, such as trade_candidate_pipeline.py, to align with individual trading strategies.