A neural search engine that enables AI agents to perform semantic web research and extract clean, LLM-ready content without scraping.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install web-search-exa
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-search-exa 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 Exa Neural Web Search & Research?
Exa is a cutting-edge neural search engine designed specifically for AI agents and large language models. Unlike traditional keyword-based search engines, Exa understands the underlying meaning and context of a query, allowing it to find relevant pages based on descriptive intent. This Openclaw Skills integration provides clean, markdown-formatted content directly to your agent, bypassing the need for manual scraping or complex HTML parsing. It transforms the web into a structured database that your agent can query naturally.
By utilizing this skill, developers can equip their agents with the ability to conduct deep research, find academic papers, and monitor real-time news with high precision. Whether you are building a competitive intelligence tool or a coding assistant, this skill ensures your agent has access to the most relevant, high-quality information available on the live web.
Exa Neural Web Search & Research Use Cases
- Perform deep, multi-step research on specific industries to generate comprehensive competitive analysis reports.
- Retrieve highly specific code examples and documentation from sources like GitHub and Stack Overflow.
- Find and extract professional profiles, bios, and expert backgrounds from LinkedIn and personal sites.
- Monitor fresh content such as news updates, tweets, or financial reports with specific date and age filters.
- Locate semantically similar pages to a provided URL to expand a research corpus.
How Exa Neural Web Search & Research Works
- The AI agent initiates a request by describing the target information in natural language rather than just keywords.
- The request is processed through the Exa neural search engine to identify semantically matching web pages.
- Search parameters are applied to filter results by domain, date range, or specific categories like research papers or news.
- The skill fetches the page content and converts it into LLM-ready Markdown, highlights, or structured JSON summaries.
- The processed data is returned to the agent, allowing it to synthesize answers with grounded citations or continue iterative research steps.
Exa Neural Web Search & Research Setup
To integrate this search capability into your workflow using Openclaw Skills, add the MCP server to your configuration with the following command:
openclaw mcp add exa --url "https://mcp.exa.ai/mcp"
To access advanced tools and remove rate limits, include your API key in the connection URL:
# Example with API Key
https://mcp.exa.ai/mcp?exaApiKey=YOUR_EXA_KEY
You can also specify which optional tools to enable by appending a comma-separated list to the tools parameter in the URL query string.
Exa Neural Web Search & Research Data Schema & Taxonomy
The skill organizes its data to be highly token-efficient for agentic workflows. You can control the output schema using the following content modes:
| Content Mode | Output Format | Purpose |
|---|---|---|
| highlights | Extractive snippets | Best for factual lookups and saving tokens |
| text | Full Markdown | Used for deep analysis of an entire page |
| summary | LLM-generated | Provides quick overviews or structured data extraction |
Advanced tools like deep_search_exa allow for a custom outputSchema using JSON Schema, enabling your agent to return structured data objects directly from search results (e.g., a list of companies with specific metadata fields).
Exa Neural Web Search & Research Advanced Features
- Supports asynchronous deep research tasks that search, read, and compile full reports over multiple steps.
- Features a specialized crawling engine that handles JavaScript-rendered pages and complex PDF layouts effectively.
- Provides granular freshness control via the
maxAgeHoursparameter for real-time data retrieval. - Allows for semantic similarity searching where a URL can be used as the query to find related content.
- Includes built-in category targeting to refine searches specifically for research papers, tweets, or financial reports.
SKILL.md
Loading
Related Openclaw Skills
Hevy CLI Skill
A powerful command-line interface for accessing and analyzing Hevy fitness tracking data including workouts, routines, and exercise templates.
Nosi Publishing Skill
A streamlined publishing utility that allows AI agents to push text and markdown to the web, returning a permanent shareable URL.
Travel Agent by BonBook
A full-service AI travel agent that enables flight discovery, booking, and management through simple email commands.
Huangli Almanac Query
A professional-grade interface for AI agents to query the traditional Chinese Almanac (Huangli) for auspicious dates and daily activity guidance.
DomainKits MCP Server
A powerful domain intelligence toolset for searching newly registered domains and performing reverse nameserver lookups through MCP-compatible AI clients.
OpenAI TTS
A professional-grade text-to-speech skill leveraging OpenAI's API to generate natural-sounding audio with advanced voice and speed controls.
Featured*
