AI Skill

Tiger OpenAPI provides an AI Skill based on the open Agent Skills standard. It supports 40+ AI coding tools, including Claude Code, Cursor, Gemini CLI, GitHub Copilot, VS Code, and OpenAI Codex, as well as frameworks such as OpenClaw.

Once installed, the plugin enables AI tools to help you retrieve market data, place orders, manage accounts, and perform other tasks through natural-language requests without manually searching the API documentation.

Installation

Install with a single command using npx skills:

npx skills add tigerfintech/tigeropen-skill

The CLI automatically detects installed AI coding tools and prompts you to choose project or global scope.

More installation options:

# List available skills
npx skills add tigerfintech/tigeropen-skill --list

# Install for a specific tool
npx skills add tigerfintech/tigeropen-skill -a claude-code

# Install globally (available across all projects)
npx skills add tigerfintech/tigeropen-skill -g

# Install to all detected tools
npx skills add tigerfintech/tigeropen-skill --all

Alternative installation methods

Load Locally with --plugin-dir (Claude Code)

git clone https://github.com/tigerfintech/tigeropen-skill.git ~/tigeropen-skill
claude --plugin-dir ~/tigeropen-skill

Copy Manually to the Skills Directory

Clone the repository first, then copy the skill directory for your SDK:

git clone https://github.com/tigerfintech/tigeropen-skill.git /tmp/tigeropen-skill

Choose the skill directory name for your language:

SDK	Skill directory
Python	`tigeropen`
Java	`tigeropen-java`
C++	`tigeropen-cpp`
C#	`tigeropen-csharp`
Go	`tigeropen-go`
Rust	`tigeropen-rust`
TypeScript	`tigeropen-typescript`

# Global (all projects) — replace "tigeropen" with your SDK's directory name
cp -r /tmp/tigeropen-skill/skills/tigeropen ~/.claude/skills/
rm -rf /tmp/tigeropen-skill

# Project-specific
mkdir -p .claude/skills
cp -r /tmp/tigeropen-skill/skills/tigeropen .claude/skills/

Prerequisites

Choose the SDK that matches the Skill language and use the SDK links under Quick Start to open that language's installation guide (C++, C#, Go, Rust, and TypeScript each have a separate quick start). Only the Python Skill requires tigeropen; for example, run uv add tigeropen (or pip install tigeropen inside an activated virtual environment)
Have a Tiger Brokers account with API access (Developer Page)
Have your tiger_id, private key file, and account ready

Skill Modules

The skill includes the following modules. The AI automatically loads the relevant reference for each request:

Module	Coverage
Quick Start	SDK setup, configuration, client creation, enum and object references, error codes, and FAQs
Market Data	Stock, option, futures, and fund market data; candlestick bars; depth; scanners; and fundamentals
Trading	Market, limit, stop, and algorithmic orders; order management; assets; positions; and fund transfers
Options	Option chains, single-leg/multi-leg combo strategies, and current-input Greek calculation tools
Real-Time Streaming	Real-time quote, depth, tick, candlestick-bar, order, position, and asset updates
CLI	CLI configuration, market data, trading, and account information
MCP Server	MCP Server setup for Cursor, Claude Code, and Trae

Usage Examples

After installation, submit natural-language requests in your AI tool:

> Get real-time quotes for AAPL and TSLA

> Buy 100 shares of AAPL with a limit order at $150

> Fetch AAPL daily candlestick bars (K-line data) for the last 30 days and plot them

> Show my account assets and current positions

> Get the AAPL option chain expiring next month, then use [Option Calculation Tools](/docs/other-example#option-calculation-tools) with current market inputs to calculate Greeks

> Subscribe to real-time quote updates for AAPL and TSLA

Note: Option-chain Greek request flags, filters/models, and response fields delta, gamma, theta, vega, and rho are deprecated. These values update daily and are not suitable for intraday use. AI tools must not use them to filter contracts or make real-time trading decisions. Use Option Calculation Tools with current market inputs instead.

How It Works

An AI Skill is a set of structured reference documents that teaches AI coding tools how to call Tiger OpenAPI correctly. When you describe a task in an AI tool:

The AI identifies your intent, such as retrieving market data or placing an order.
It loads the relevant reference document, such as quote.md or trade.md.
It generates code from the API specifications and examples in that document.
It executes the code and returns the results.

You do not need to memorize API details; describe the task in natural language instead.

Supported AI Tools

The skill follows the open Agent Skills standard and supports the following AI coding tools, among others:

Claude Code — Anthropic's official CLI
Cursor — AI code editor
Gemini CLI — Google Gemini command-line tool
GitHub Copilot — GitHub AI programming assistant
VS Code — Supported via extensions
OpenAI Codex — OpenAI coding tool
Kiro — AWS AI IDE
Trae — ByteDance AI IDE

It also supports 40+ other tools that implement the Agent Skills standard.

Repository

GitHub: tigerfintech/tigeropen-skill

Disclaimer

This AI Skill provides technical reference material that helps AI coding tools call Tiger OpenAPI. It does not constitute investment advice. AI-generated code, strategies, or trading instructions may contain errors or omissions. Independently review and validate all output before executing any trade. We make no warranty regarding the accuracy, completeness, or suitability of AI-generated content. You are responsible for protecting your API keys and trading credentials. You assume all risks and losses arising from use of this tool, including losses caused by AI output errors, network latency, or credential exposure. The tool is provided "as is" without warranties of any kind. For the full terms, see the Risk Disclosure and Disclaimer.