anysearch-ai/anysearch-skill

What it solves

AnySearch Skill provides AI agents with a unified interface for real-time web search and content extraction. It eliminates the need for agents to manage multiple search providers or handle complex HTML parsing, providing a streamlined way to retrieve current information from the general web or specific vertical domains.

How it works

The project provides a set of Command Line Interface (CLI) scripts in multiple languages (Python, Node.js, PowerShell, and Bash) that act as a wrapper around the AnySearch API. AI agents can install the skill and use these scripts to perform the following actions:

• General and Vertical Search: Performing searches across the web or within specific domains.
• Batch Search: Executing multiple search queries in parallel to gather information more efficiently.
• Parallel Batch Search: Using the batch_search command to handle multiple queries at once.
• Content Extraction: Using the extract command to convert full-page web content into a clean, agent-friendly Markdown format.

Who it’s for

This tool is designed for developers building AI agents (such as those using Claude Code, Cursor, Windsurf, or OpenCode) and agent platforms that support a skill marketplace.

Highlights

• Multi-Runtime Support: Works across Python, Node.js, Shell, and PowerShell, ensuring compatibility with various agent environments.
• C-API Integration: Simple API key registration process that allows agents to automatically register users via email without manual signup.
• Markdown Extraction: Automatically converts web pages into Markdown, simplifying the process for LLMs to process external data.
• Flexible Configuration: Supports API keys via CLI flags, .env files, or environment variables, with a fallback to anonymous access.