An advanced AI-driven financial analysis skill designed to detect institutional accumulation in A-share AI stocks using a multi-factor 5-dimensional scoring engine.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install institutional-tracker-ai
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install institutional-tracker-ai using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Institutional Tracker AI?
Institutional Tracker AI is a high-performance workflow optimized for Openclaw Skills that identifies institutional buying behaviors within the A-share AI industry chain, including computing power, semiconductors, and large language models. By leveraging Tushare structured data, the skill synthesizes five critical signal dimensions—fund flow, volume-price dynamics, chip distribution, north-bound capital, and event-driven volatility—to generate statistically significant buy signals with a historical 10-day win rate of over 55%.
The skill is built with risk management at its core, featuring an automated market regime detection system that suppresses buy signals during bear markets. It serves as a professional-grade tool for developers and quantitative analysts looking to automate market scanning and backtesting within the Openclaw Skills ecosystem.
Institutional Tracker AI Use Cases
- Automating daily post-market scans of 25+ leading AI sector stocks to identify institutional accumulation patterns.
- Performing large-scale backtesting (2022-2026) to validate trading algorithms across different market cycles.
- Dynamically adjusting trade entry thresholds based on macro market regimes like Bull, Bear, or Neutral states.
- Utilizing Level-2 minute-line data to recognize intraday VWAP-based institutional accumulation features.
How Institutional Tracker AI Works
- Data Ingestion: The system fetches daily, money flow, and minute-level data via the Tushare API using a local secure token.
- 5-Dimensional Scoring: It applies a weighted algorithm to evaluate Fund Flow (30%), Volume-Price (24%), Chip Position (18%), North-bound/Margin (16%), and Events (12%).
- Market Regime Analysis: An 11-indicator macro check determines the market state; in a Bear regime, the skill sets entry thresholds to a prohibitive level to protect capital.
- Intraday Validation: The skill analyzes 5-minute candles for tail-end volume spikes and consistent VWAP behavior to confirm institutional presence.
- Signal Output: The workflow generates a final status for each ticker, such as Accumulation (Buy), Markup (Hold), or Washout (Add).
Institutional Tracker AI Setup
To deploy this skill within your Openclaw Skills environment, ensure you have Python 3.10+ and a valid Tushare Token. Follow these steps:
# Copy the skill scripts to your local workspace
cp -r ~/.workbuddy/skills/institutional-tracker-ai/scripts/ ./institutional_tracker/
# Configure your local credentials
vi ./institutional_tracker/config.py
# Set TUSHARE_TOKEN = "your_token_here"
# Create the necessary data and report directories
mkdir -p ./institutional_tracker/data/daily_scores
mkdir -p ./institutional_tracker/reports
# Execute the daily scan
cd ./institutional_tracker && python3 main.py
Institutional Tracker AI Data Schema & Taxonomy
The skill manages its data through a structured hierarchy to ensure transparency and ease of integration:
| Component | Storage Path | Description |
|---|---|---|
| Daily Scores | /data/daily_scores/ |
Historical factor scores and daily signals. |
| HTML Reports | /reports/ |
Visual summaries of scan results for human review. |
| Factor Weights | main.py |
Configurable weights for the 5-dimension scoring engine. |
| Market State | config.py |
Cached market regime status (Bull/Bear/Neutral). |
Data is processed locally, ensuring that sensitive API tokens and custom stock pools remain within your private Openclaw Skills environment.
Institutional Tracker AI Advanced Features
- Multi-period backtesting framework supporting training (2022-2024) and out-of-sample testing (2025-2026) datasets.
- Automated WeChat notification hooks for pushing daily HTML report summaries and signal alerts.
- Statistical significance testing using P-values and Wilson confidence intervals to prevent over-fitting.
- Integrated retail sentiment filter that penalizes FOMO-driven price action and rewards quiet institutional accumulation.
SKILL.md
Loading
Related Openclaw Skills
Customer Success Manager
An expert AI agent skill designed to manage customer lifecycles, drive product adoption, and maximize lifetime value through data-driven retention and growth strategies.
Polymarket Politics Random Buyer
An automated AION trading template designed to discover politics-related Polymarket events and execute trades with built-in safety guardrails.
C# .NET Developer Skill
An expert AI agent skill for building high-performance .NET 8+ applications, cloud-native solutions, and modern C# architectures.
Kaiber AI
A high-performance AI video generation skill that transforms static images and raw clips into professional 1080p animated videos using cloud-based GPU rendering.
Eaton Industrial Insights
A specialized skill for researching Eaton's transition from automotive parts to a global leader in electrification and data center power management.
Agent Self-Evolution
A structured framework for AI agents to safely improve their own prompts, skills, and memory through architecture-first evolution.
Featured*
