An asynchronous Python-based tool that enables AI agents to aggregate and deduplicate search results from multiple search engines simultaneously.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install web-multi-search
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 web-multi-search 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 Web Multi-Search?
Web Multi-Search is a high-performance search utility designed to broaden the information gathering capabilities of AI agents. By utilizing the async-search-scraper library, this skill allows agents to query Bing, Yahoo, Startpage, Aol, and Ask in a single operation. It is specifically engineered to handle the complexities of web scraping across different platforms, making it a vital component for those building advanced Openclaw Skills.
The tool is optimized for speed and reliability, supporting asynchronous execution to ensure that data retrieval doesn't become a bottleneck. It provides a structured way for agents to access the open web, bypassing the limitations of single-provider searches and providing a more comprehensive data set for analysis, summary, or fact-checking tasks.
Web Multi-Search Use Cases
- Gathering diverse perspectives by aggregating results from multiple search providers.
- Performing deep-dive research by iterating through multiple result pages automatically.
- Privacy-conscious searching through the use of SOCKS5 or HTTP proxies.
- Streamlining data pipelines by deduplicating URLs and domains at the source.
- Providing real-time context to LLMs for up-to-date query resolution.
How Web Multi-Search Works
- The AI agent or user executes the search script with a specific query and optional flags for engines or page depth.
- The skill initiates asynchronous HTTP requests to the selected search engines (e.g., Bing, Yahoo, Ask).
- It iterates through the specified number of result pages for each engine to maximize data coverage.
- Result snippets, titles, and URLs are parsed and structured into a unified data format.
- Optional deduplication logic is applied to remove redundant links or domains.
- The final structured dataset is returned as JSON, CSV, or plain text for the agent to process.
Web Multi-Search Setup
To integrate this into your environment, navigate to the skill directory and install the necessary dependencies. Note that the scraper library must be installed directly from the source repository.
cd skills/web-multi-search
python3 -m pip install -r requirements.txt
python3 -m pip install git+https://github.com/soxoj/async-search-scraper.git --no-deps
For Linux users without pip access, ensure system packages are updated:
sudo apt-get update
sudo apt-get install -y python3-requests python3-aiohttp python3-aiohttp-socks python3-bs4
Web Multi-Search Data Schema & Taxonomy
The skill returns a structured array of search result objects. The default JSON output follows this schema:
| Attribute | Type | Description |
|---|---|---|
| engine | string | The source engine (e.g., Bing, Yahoo) |
| host | string | The domain hosting the result |
| link | string | The full destination URL |
| title | string | The metadata title of the page |
| text | string | A descriptive snippet of the page content |
Web Multi-Search Advanced Features
- Support for concurrent engine querying to reduce latency.
- Built-in deduplication filters for unique-urls and unique-domains to clean output.
- Proxy integration (HTTP/SOCKS5) for rotating identities and bypassing rate limits.
- Multiple output formats including JSON, CSV, and text for flexible integration into other Openclaw Skills.
- Deep pagination support to retrieve results beyond the first page of search engines.
SKILL.md
Loading
Related Openclaw Skills
OSINT Graph Analyzer
A powerful tool for building Neo4j knowledge graphs from OSINT data to discover hidden connections and community patterns.
MCP Registry Manager
A centralized tool for discovering, scoring, and managing Model Context Protocol (MCP) servers through semantic search and quality metrics.
Agentic Compass
Agentic Compass is a local-only self-reflection tool that transforms AI agent memory logs into objective action plans and measurable performance metrics.
Agent Trust Validator
A unified verification layer that validates AI agent identities and reputation scores across multiple on-chain and off-chain trust protocols.
Workspace Review
A comprehensive self-audit tool designed to verify workspace file structures and ensure adherence to standard OpenClaw conventions.
Polymarket FastLoop Trader (Improved)
An advanced algorithmic trading skill for Polymarket short-term markets using CEX momentum and multi-signal confirmation.
Featured*
