MySearch for Openclaw

MySearch is a high-performance search aggregator that routes queries to Tavily, Firecrawl, Exa, and X through a unified interface.

skernelx
v0.1.11
Mar 20, 2026
1
1k
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install mysearch

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install mysearch 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 MySearch?

MySearch is the default search skill for OpenClaw, acting as a powerful multi-provider aggregator designed to replace legacy, single-engine tools. It provides a consolidated entry point for any external lookup, whether you need real-time web results, technical documentation, or social media trends. By leveraging this skill within the ecosystem of Openclaw Skills, developers can ensure their agents always access the most relevant data source for a given task.

Technically, MySearch intelligently routes requests based on task type: Tavily for general search, Firecrawl for technical docs and GitHub data, and xAI for social sentiment. This strategic routing minimizes latency and maximizes context quality. It supports a streamlined proxy architecture, allowing you to manage multiple provider credentials through a single base URL and API key, significantly reducing the complexity of managing your Openclaw Skills environment.

MySearch Use Cases

  • Aggregating real-time news and web results for general AI assistant queries.
  • Scraping and extracting markdown content from documentation, pricing pages, and changelogs.
  • Monitoring social media trends and public discussions on X (Twitter).
  • Performing deep technical research across GitHub repositories and PDF documents.
  • Verifying facts by comparing news reports against social media activity using multi-source strategies.

How MySearch Works

  1. The AI agent triggers a search or extract command via the mysearch_openclaw.py script.
  2. The skill identifies the search mode (web, social, news, etc.) and analyzes the user's intent.
  3. Requests are routed to the primary provider, such as Tavily for web or Firecrawl for documentation.
  4. If a primary provider returns insufficient data, MySearch utilizes fallback providers like Exa to fill the gaps.
  5. The aggregated data is cleaned, structured into markdown, and returned with a clear distinction between facts and source URLs.

MySearch Setup

To integrate this search engine into your collection of Openclaw Skills, follow these steps:

# Install the skill to your local OpenClaw directory
bash scripts/install_openclaw_skill.sh --install-to ~/.openclaw/skills/mysearch

# Configure your environment variables in openclaw.json
# Example config:
# "mysearch": {
#   "env": {
#     "MYSEARCH_PROXY_BASE_URL": "https://your-proxy.com",
#     "MYSEARCH_PROXY_API_KEY": "your-key"
#   }
# }

# Verify the installation and provider health
python3 scripts/mysearch_openclaw.py health

MySearch Data Schema & Taxonomy

MySearch organizes data dynamically based on the requested mode. The following parameters define the schema for queries:

Parameter Description Allowed Values
query The search string or URL to extract Any string
mode The search domain focus web, news, social, docs, research, github, pdf, auto
intent The logical goal of the search status, resource, exploratory
sources Specific providers to query web, x, news

The output is returned as a structured markdown response prioritizing a summary conclusion followed by source URLs and metadata tags.

MySearch Advanced Features

  • Unified Proxy Architecture: Manage Tavily, Firecrawl, and Exa using a single set of proxy credentials.
  • Intelligent Intent Routing: Automatically selects the best search engine based on whether the user needs news, code, or social trends.
  • Multi-Source Verification: Compare results across web and social sources in a single research execution.
  • Fallback Extraction: Automatically falls back to Tavily extract if Firecrawl fails to parse a specific URL.
  • High-Density Context: Optimized for Openclaw Skills to provide the most token-efficient results for LLM consumption.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*