An autonomous web research skill that enables AI agents to search, browse, and extract real-time knowledge from the internet.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install web-learner
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-learner 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 Learner?
The Web Learner skill transforms an AI agent into a proactive researcher capable of navigating the live internet. By integrating specialized tools for search, fetching, and full browser interaction, it allows the agent to transcend its training data limits. This ensures that the agent can provide current news, localized weather, financial data, and deep-dive technical documentation from any public URL. As one of the versatile Openclaw Skills, it bridges the gap between static LLM knowledge and the ever-changing web.
This skill is particularly effective for users who need their AI to stay updated with the latest events or perform deep-dives into topics that require real-time verification. It manages the complexity of web navigation, from simple keyword searches to handling JavaScript-heavy pages via a full browser service.
Web Learner Use Cases
- Monitoring latest news and global events in real-time.
- Learning and summarizing new technical topics or documentation from the web.
- Fetching live data such as weather forecasts, stock prices, or sports scores.
- Extracting and converting webpage content into clean Markdown for data processing.
- Verifying facts or supplementing information not found in the agent's internal training data.
How Web Learner Works
- Demand Analysis: The agent identifies the specific information required based on user intent and determines if web access is necessary.
- Tool Selection: It selects the best tool for the job: web_search for broad discovery, web_fetch for direct content extraction, or the browser tool for interactive sites.
- Information Retrieval: The skill executes the search or crawl, utilizing parameters like region (e.g., CN) and freshness (e.g., daily or weekly) to ensure relevance.
- Integration and Processing: Raw data is filtered to extract key insights, handle potential anti-crawler obstacles, and convert content into readable formats.
- User Delivery: The synthesized information is presented to the user in structured Chinese, complete with source citations for transparency.
Web Learner Setup
To enable this skill within your Openclaw Skills library, you must configure the necessary API keys and services. Ensure your environment has access to the Brave Search API and a running browser service for advanced rendering.
# Example configuration for web search capabilities
export BRAVE_API_KEY='your_api_key_here'
# Ensure the browser service is active for JS-heavy sites
# Follow your specific AI agent's instructions to link the browser tool
Web Learner Data Schema & Taxonomy
| Data Type | Description | Format |
|---|---|---|
| Search Results | Metadata, links, and snippets retrieved from search engines | JSON |
| Web Content | Cleaned text or Markdown extracted from target URLs | Markdown |
| Visual Data | Screenshots of web interfaces for visual context | PNG |
| Source Logs | Records of visited URLs and timestamps for attribution | Table |
Web Learner Advanced Features
- Regional optimization with country-specific search parameters to prioritize local results.
- Time-sensitive filtering using freshness constraints to get results from today, this week, or this month.
- Automated fallback logic that switches to full browser rendering if standard fetching fails due to JavaScript requirements.
- Intelligent extraction of video descriptions and document metadata when direct playback or download is restricted.
- Seamless multi-source synthesis designed to work alongside other Openclaw Skills for comprehensive reporting.
SKILL.md
Loading
Related Openclaw Skills
Video Understanding Skill
A powerful AI skill for extracting metadata, summarizing content, and identifying key timestamps from global video platforms like YouTube and Bilibili.
Email Reader
An AI-driven email management skill that enables agents to read, summarize, and send messages through IMAP and SMTP protocols using the himalaya CLI.
Document Pro
A professional-grade document processing skill that enables AI agents to read, parse, and extract key information from PDF, Word, and PowerPoint files.
flightclaw
A comprehensive flight price tracking and search tool that integrates Google Flights data into your local environment and AI agents.
Creating R Research Projects
A comprehensive skill for automating reproducible R-based research workflows, from data cleaning to publication-ready reports.
Debugging R Environment and Dependencies
A comprehensive diagnostic and repair tool for resolving R environment issues, package conflicts, and system dependency failures.
Featured*
