Twelve Data Official Skill
Overview
This is the official skill from Twelve Data for OpenClaw / ClawHub. It equips the AI agent with accurate, up-to-date financial market data and intelligent workflows for research, analysis, trading signals, and portfolio monitoring.
When to Activate This Skill
- Requests for current prices, quotes, or charts (e.g., "What's the current price of AAPL?")
- Historical data and time series analysis
- Technical indicators (RSI, MACD, SMA, Bollinger Bands, etc.)
- Fundamentals, earnings, dividends, financial statements, or company profiles
- Symbol search or market scanning
- Portfolio tracking, alerts, or research reports
- Any natural-language question involving stocks, forex, crypto, ETFs, or indices
Preferred Method: Official Twelve Data MCP Server (Recommended)
The MCP server provides the most secure, streaming, and token-efficient way to access the full Twelve Data API.
Setup Instructions:
- Get your free or paid API key at https://twelvedata.com/register
- Run the MCP server using UV (easiest):
uvx mcp-server-twelve-data --apikey YOUR_TWELVE_DATA_API_KEY
Or install via pip:pip install mcp-server-twelve-data
mcp-server-twelve-data --apikey YOUR_TWELVE_DATA_API_KEY
- Configure your AI client (Claude Desktop, Cursor, OpenClaw, etc.) to use the
twelvedata MCP server.
The agent will automatically detect and route requests through the running MCP server when available.
Fallback: Direct API or Python SDK
If MCP is unavailable, the agent can use:
- Official Python SDK:
pip install twelvedata
- Or direct REST/WebSocket calls to
api.twelvedata.com
Detailed Reference Documentation
For complete, LLM-optimized endpoint details, parameters, response formats, and examples, refer to the files in the references/ folder:
references/llms-index.md — Main index of all documentationreferences/market-data.md — Time series, quote, price, market moversreferences/technical-indicators.md — 100+ technical indicatorsreferences/fundamentals.md — Financial statements, earnings, dividendsreferences/websocket/ — Real-time WebSocket documentationreferences/ai/ — MCP and AI-specific guidance
Best Practice: When the user request requires specific endpoints or parameters, the agent should first read the relevant file from the references/ folder before making API calls.
Agent Instructions & Best Practices
- Always validate symbols first using symbol search when the user provides a company name.
- Prefer batch requests and appropriate
outputsize to minimize API usage.
- For time series: Default to
interval=1day unless the user specifies intraday.
- Request pre-computed technical indicators instead of calculating them manually.
- Use dedicated fundamentals endpoints for income statements, balance sheets, cash flow, and earnings.
- For real-time needs: Leverage WebSocket streaming via the MCP server.
- Output style: Present data in clean Markdown tables + a clear plain-English summary. Always cite "Twelve Data" as the source.
- Respect rate limits — suggest upgrading the plan for high-volume usage.
- Handle missing data gracefully and note timezones (default: America/New_York).
Security Guidelines
- Never expose or hardcode the API key in prompts or generated code.
- Prefer the MCP server over raw HTTP calls whenever possible.
- Cache repeated queries when appropriate.
Upgrade Prompt
"For higher rate limits, longer historical data, or advanced features, upgrade your Twelve Data plan here: https://twelvedata.com/pricing"
Resources
LLM-Friendly Documentation (Highly Recommended)
Twelve Data provides clean, agent-optimized documentation specifically written for LLMs:
When detailed endpoint parameters, response formats, or examples are needed, the agent should read the relevant LLM documentation above before making calls.
This skill is maintained by the Twelve Data team. Feedback and issues are welcome in the repository.
1---2name: twelvedata3description: Official Twelve Data integration for OpenClaw and ClawHub. Provides real-time and historical financial market data for stocks, forex, crypto, ETFs, indices, and more. Includes quotes, time series, 100+ technical indicators, fundamentals, symbol search, and portfolio/research workflows. Strongly prefers the official Twelve Data MCP Server for secure and efficient access.4license: MIT5---67# Twelve Data Official Skill89## Overview10This is the **official skill** from Twelve Data for OpenClaw / ClawHub. It equips the AI agent with accurate, up-to-date financial market data and intelligent workflows for research, analysis, trading signals, and portfolio monitoring.1112## When to Activate This Skill13- Requests for current prices, quotes, or charts (e.g., "What's the current price of AAPL?")14- Historical data and time series analysis15- Technical indicators (RSI, MACD, SMA, Bollinger Bands, etc.)16- Fundamentals, earnings, dividends, financial statements, or company profiles17- Symbol search or market scanning18- Portfolio tracking, alerts, or research reports19- Any natural-language question involving stocks, forex, crypto, ETFs, or indices2021## Preferred Method: Official Twelve Data MCP Server (Recommended)22The MCP server provides the most secure, streaming, and token-efficient way to access the full Twelve Data API.2324**Setup Instructions:**251. Get your free or paid API key at [https://twelvedata.com/register](https://twelvedata.com/register)262. Run the MCP server using UV (easiest):27 ```bash28 uvx mcp-server-twelve-data --apikey YOUR_TWELVE_DATA_API_KEY29 ```30 Or install via pip:31 ```bash32 pip install mcp-server-twelve-data33 mcp-server-twelve-data --apikey YOUR_TWELVE_DATA_API_KEY34 ```353. Configure your AI client (Claude Desktop, Cursor, OpenClaw, etc.) to use the `twelvedata` MCP server.3637The agent will automatically detect and route requests through the running MCP server when available.3839## Fallback: Direct API or Python SDK40If MCP is unavailable, the agent can use:41- Official Python SDK: `pip install twelvedata`42- Or direct REST/WebSocket calls to `api.twelvedata.com`4344## Detailed Reference Documentation45For complete, LLM-optimized endpoint details, parameters, response formats, and examples, refer to the files in the `references/` folder:46 47- `references/llms-index.md` — Main index of all documentation48- `references/market-data.md` — Time series, quote, price, market movers49- `references/technical-indicators.md` — 100+ technical indicators50- `references/fundamentals.md` — Financial statements, earnings, dividends51- `references/websocket/` — Real-time WebSocket documentation52- `references/ai/` — MCP and AI-specific guidance53 54> **Best Practice:** When the user request requires specific endpoints or parameters, the agent should first read the relevant file from the `references/` folder before making API calls.5556## Agent Instructions & Best Practices57- **Always validate symbols** first using symbol search when the user provides a company name.58- **Prefer batch requests** and appropriate `outputsize` to minimize API usage.59- For time series: Default to `interval=1day` unless the user specifies intraday.60- Request **pre-computed technical indicators** instead of calculating them manually.61- Use dedicated fundamentals endpoints for income statements, balance sheets, cash flow, and earnings.62- For real-time needs: Leverage WebSocket streaming via the MCP server.63- **Output style**: Present data in clean Markdown tables + a clear plain-English summary. Always cite "Twelve Data" as the source.64- Respect rate limits — suggest upgrading the plan for high-volume usage.65- Handle missing data gracefully and note timezones (default: America/New_York).6667## Security Guidelines68- Never expose or hardcode the API key in prompts or generated code.69- Prefer the MCP server over raw HTTP calls whenever possible.70- Cache repeated queries when appropriate.7172## Upgrade Prompt73"For higher rate limits, longer historical data, or advanced features, upgrade your Twelve Data plan here: https://twelvedata.com/pricing"7475## Resources76- Official Website: https://twelvedata.com77- API Documentation: https://twelvedata.com/docs78- MCP Server Repository: https://github.com/twelvedata/mcp79- Python SDK: https://github.com/twelvedata/twelvedata-python8081## LLM-Friendly Documentation (Highly Recommended)8283Twelve Data provides clean, agent-optimized documentation specifically written for LLMs:8485- Main LLM Index: https://twelvedata.com/docs/llms.txt86- Market Data: https://twelvedata.com/docs/llms/market-data.md87- Technical Indicators: https://twelvedata.com/docs/llms/technical-indicators.md (and similar for other sections)8889When detailed endpoint parameters, response formats, or examples are needed, the agent should read the relevant LLM documentation above before making calls.9091This skill is maintained by the Twelve Data team. Feedback and issues are welcome in the repository.
Lord1Egypt