ToolRegistry Hub Documentation

中文版 | English version

Welcome to the ToolRegistry Hub documentation! This document provides detailed descriptions of all tools in the project.

Ecosystem

Package	Description	PyPI	Docs
toolregistry	Core library — tool registration, schema generation, execution		Docs
toolregistry-server	Server adapters — expose tools via OpenAPI & MCP		Docs
toolregistry-hub	Ready-to-use tools — calculator, web search, file ops, etc.		Docs

📚 Documentation

For detailed documentation, please visit our ReadTheDocs pages:

• English Documentation: https://toolregistry-hub.readthedocs.io/en/latest/
• 中文文档: https://toolregistry-hub.readthedocs.io/zh-cn/latest/

Tools Overview

ToolRegistry Hub is a Python library that provides various utility tools designed to support common tasks. Here are the main tool categories:

• Calculator Tools - Provides various mathematical calculation functions
• Date Time Tools - Handles date, time, and timezone conversions
• File Operations Tools - Provides file content manipulation functions
• File System Tools - Provides file system operation functions
• Web Search Tools - Provides web search functionality
• Unit Converter Tools - Provides conversions between various units
• Other Tools - Other utility tools
• Server Mode - REST API and MCP server
• Docker Deployment - Docker containerization deployment

For detailed information about each tool category, please refer to the online documentation.

Quick Start

To use ToolRegistry Hub, first install the library:

pip install toolregistry-hub

Then, you can import and use the required tools:

from toolregistry_hub import Calculator, DateTime, FileOps

# Use calculator
result = Calculator.evaluate("2 + 2 * 3")
print(result)  # Output: 8

# Get current time
current_time = DateTime.now()
print(current_time)

Documentation Structure

This documentation is organized by tool categories, with each tool category having its own page that details all tools, methods, and usage examples under that category.

Changelog

Web Fetch Tool — Major Update (PR #140)

Structured return value

fetch_content now returns a structured dict instead of a plain str:

{
  "content": "...",
  "url": "https://example.com",
  "strategy": "readability",
  "quality": "high",
  "content_type": "text/html",
  "cached": false,
  "elapsed_ms": 123,
  "metadata": {"readability_score": 174.3, "content_length": 12345}
}

strategy parameter

A new optional strategy parameter lets you specify the extraction strategy (default: "auto"). Available choices:

Available choices are narrowed at runtime: veilrender and cdp appear only when their respective endpoints are configured.

Updated fallback chain

markdown → readability → soup → veilrender → cdp → jina → local_fallback

VeilRender fallback

VeilRender is a new optional remote headless browser fallback for JS-heavy pages and SPAs. Configure via environment variables:

VEILRENDER_ENDPOINT=https://your-veilrender-instance
VEILRENDER_TOKEN=your_token_here  # optional

Contributing

If you want to contribute to ToolRegistry Hub, please refer to the GitHub repository.