Dragon-Tiger List Deep Analyzer
Identify hot-money seats, gauge institutional vs. hot-money battles, and spot the leading stock in the sector.
The skill reads public market data but data flow relies on external APIs and third-party scripts without explicit least-privilege or user-confirmation mechanisms, and lacks data retention/privacy handling. As a financial analysis tool, data-flow transparency is moderate but sensitive-data handling and rollback are incomplete.
The skill description is clear, but no reproducible test or verification path is provided. It depends on external data sources (akshare, yfinance) and may fail on network unavailability; failure feedback is unclear. Static review cannot verify script functionality, so direct reliability evidence is insufficient.
Target scenario is clear (A-share LHB analysis), audience is explicit, but non-fit boundaries (e.g., US/HK stocks) are not declared, and reachability of external sources from mainland China is not verified. Trigger conditions match, but environment-fit evidence is limited.
Structure is clear with version and author, but lacks detailed parameter documentation and troubleshooting. README provides installation/usage examples but not lhb-analyzer-specific configuration or limitations. License is explicit (MIT), but maintenance responsibility and update path are unclear.
The core task (identifying hot-money seats) is not fully implemented in the current documentation; no directly usable output examples are provided. Dependence on external data and model inference means actual effectiveness is unverified statically, and comparative-benefit evidence is weak.
Key claims (e.g., seat matching) rely on lib/seat_db.py, but static review cannot reproduce them, and there is no independent verification. External data dependencies and update frequency are unclear, limiting traceability.
- The skill relies on external data sources (e.g., akshare, yfinance) whose reachability from mainland China is not verified, potentially affecting usability.
- Static review cannot verify actual script functionality; test in a real environment before use.
- No user-confirmation mechanism or data-privacy statement is provided; handle sensitive information cautiously.
What does this skill do, and when should you use it?
This is an A-share market Dragon-Tiger List (LHB) analysis skill. It fetches raw LHB data via Python scripts, matches hot-money seats using a built-in seat database, and assesses the buying and selling landscape. The output includes hot-money identification, institutional vs. hot-money comparison, sector leader ranking, and a concluding remark. Developed by FloatFu-true and released under MIT license, this skill is part of the UZI-Skill monorepo.
The skill accepts a stock ticker or name, runs scripts/fetch_lhb.py to retrieve LHB data, uses lib/seat_db.py's match_seats_in_lhb() and is_in_range() to identify hot-money seats and determine if they are within range. It also pulls sector-wide LHB data for comparison. It generates a Markdown report with a list of identified hot-money seats, institutional vs. hot-money comparison, sector leader ranking, and a one-sentence conclusion.
- Check who is buying and selling a specific stock, especially hot-money seat activity.
- Determine whether a stock is institutionally or hot-money dominated to inform investment decisions.
- Compare LHB data across stocks in the same sector to identify a high-recognition leader.
- Track whether specific hot-money traders (e.g., Zhang Mengzhu, Foshan Wuyingjiao) appear on the list and their trading direction.
- Quickly review a stock's recent LHB appearance frequency and capital flows.
What are this skill's strengths and limitations?
- Free data sources, no API keys required
- Built-in encyclopedia of 22 hot-money seats for accurate identification
- Provides institutional vs. hot-money comparison to reveal market dynamics
- Automatically compares sector peers to aid leader discovery
- Only covers A-share LHB, not HK or US stocks
- Depends on public data sources, which may be affected by network restrictions
- Skill is maintained by a single developer (FloatFu-true), potentially lacking sustained updates
- Evaluation relies on a simplified seat database that may not cover all hot-money players
How do you install this skill?
This skill is one of five in the UZI-Skill repository. Clone the whole repository and install Python dependencies. The skill lives in skills/lhb-analyzer/; ensure that directory contains SKILL.md, scripts, and reference files.
How do you use this skill?
In any Agent Skills-compatible client (e.g., Claude Code, Codex), ask directly about LHB analysis, e.g., 'Analyze the LHB for 600519'. The skill will run the scripts and return a formatted analysis.
How does this skill compare with similar options?
This skill is part of the UZI-Skill collection; the repo also includes deep-analysis, investor-panel, etc., but this skill focuses specifically on LHB analysis.