Financial Report Minesweeper
Automated A-share financial fraud and risk detection tool based on Tang Chao's methodology.
Automate A-share financial fraud and risk detection using 30 structured checklist rules.
This tool automates Tang Chao's financial report analysis methodology into a set of checks for Claude Code. It suits accountants, auditors, and equity analysts reviewing A-share companies against 30 rules across 7 layers.
What it is
It fetches 10 years of structured financial data via Tushare API and downloads annual report PDFs to execute a 7-layer checklist. It flags anomalies like inventory-margin divergence, cash flow discrepancies, and non-standard audit opinions before outputting a risk-rated report.
Who it's for
It suits financial analysts and investors who use Claude Code and need a programmatic sanity check on A-share company filings. It does not suit users outside the A-share market or those who do not use the Claude Code CLI environment.
Key features
7-Layer Rule Checklist
Evaluates 30 programmatic risk rules spanning audits, profits, cash flows, and cross-table validation.
Tushare API Integration
Pulls 10 years of structured balance sheets, income statements, cash flows, and peer comparisons automatically.
PDF Qualitative Parsing
Downloads and inspects annual report PDFs for non-structured footnotes, management reports, and transport fees.
Local Output Generation
Saves raw data, execution logs, and final risk-rated markdown reports directly into an output directory.
Up and running in 3 steps
Our honest verdict
What's great
- Codifies Tang Chao's established financial fraud detection book methodology into automated scripts
- Performs cross-table validation between balance sheets, income statements, and cash flow data
- Automates data retrieval from Tushare and annual report PDF downloads without manual entry
- Flags high-risk anomalies such as inventory turnover drops paired with rising gross margins
Where it falls short
- Requires an active Claude Code CLI environment and a registered Tushare Pro token to function
- Focused exclusively on Chinese A-share market data structures and accounting rules
- Relies on external API availability and limits for data ingestion
Pricing & license
The tool is open source with no licence fees mentioned in the repository. The only operational cost is maintaining your Python environment and any usage tiers associated with your Tushare Pro token.
Open remote AI-training roles
873 open remote AI-training roles, updated November 15, 2026; most pay $20–$170 an hour. Zekai lists them free; the application goes to the employer’s platform.
Alternatives
These reviewed AI tools do the same job instantly — from our scored directory.
Use it, or have us set it up
FAQ
What data sources does the tool use?
It combines Tushare API for 10 years of structured quantitative data with direct annual report PDF downloads for qualitative footnote analysis.
Does this replace reading the original book?
No. The author and maintainers strongly recommend reading Tang Chao's book to understand the underlying framework before using the tool.
What environment is required to run it?
You need Python 3.9+, a Tushare Pro token, and the Claude Code CLI installed on your system.
