A high-speed Google search integration that extracts full-text web content for comprehensive AI research.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install serper
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 serper 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 Serper Search & Extraction?
The Serper skill is a robust search and retrieval tool within the Openclaw Skills ecosystem designed to provide AI agents with real-time, deep access to web data. It distinguishes itself by going beyond simple metadata or short search snippets; instead, it utilizes the Serper API to identify relevant sources and then concurrently scrapes those pages to extract clean, full-text content using the trafilatura library. This ensures that agents have the necessary depth of information for complex reasoning, fact-checking, and detailed synthesis.
By integrating this capability, developers can build agents that understand the full context of a webpage, from technical documentation to long-form news articles. These Openclaw Skills are optimized for performance, employing parallel processing and strict 3-second timeouts to maintain a highly responsive user experience even when fetching multiple results.
Serper Search & Extraction Use Cases
- Conducting deep research on technical topics that require full-text analysis rather than short snippets.
- Tracking current events and breaking news through specialized time-sensitive search modes.
- Comparing products, prices, and reviews across multiple live web sources.
- Performing localized searches in non-English languages with precise country and language targeting.
- Automating factual verification for AI agents by providing source-grounded full article text.
How Serper Search & Extraction Works
- The skill initiates a fast Google search via the Serper API to retrieve a set of relevant URLs based on the query.
- It immediately triggers concurrent page scraping for all result URLs using a high-performance extraction engine.
- Each page is processed with a strict 3-second timeout to prevent bottlenecks and ensure rapid data delivery.
- The extracted content is cleaned of boilerplate and streamed back as a JSON array, allowing the agent to begin processing the first results while others are still being fetched.
Serper Search & Extraction Setup
To use this component of the Openclaw Skills library, you must have a Serper API key configured in your environment. Use the following command structure to invoke the skill:
# Standard English search
python3 scripts/search.py -q "your search query"
# Recent news search with locale flags
python3 scripts/search.py -q "aktuelle Nachrichten" --mode current --gl de --hl de
Serper Search & Extraction Data Schema & Taxonomy
The skill returns a streamed JSON array containing search metadata followed by individual result objects. The data is organized as follows:
| Field | Description |
|---|---|
title |
The title of the web page or news article. |
url |
The direct source URL of the result. |
source |
Classification of the result (web, news, or knowledge_graph). |
content |
The full extracted text of the page, falling back to a snippet if extraction fails. |
date |
Timestamp or relative date for news and time-sensitive web results. |
locale |
Metadata regarding the country and language codes used for the search. |
Serper Search & Extraction Advanced Features
- Two distinct operational modes:
defaultfor evergreen research andcurrentfor recent news and past-week developments. - Multi-threaded parallel scraping that extracts full page content from up to 6 sources simultaneously.
- Granular locale control using
--gl(country) and--hl(language) flags for accurate international results. - Streamed output architecture that minimizes time-to-first-byte for agent processing.
- Built-in query discipline guidelines to optimize API usage and reduce redundant web traffic.
SKILL.md
Loading
Related Openclaw Skills
Openclaw Serper Search & Extraction
A high-performance search skill that retrieves Google results and extracts full-page readable text for deep context research.
Openclaw Feeds: RSS News Aggregator
A lightweight RSS aggregator that fetches real-time headlines across news, gaming, and finance categories without requiring external API keys.
Drafts CLI
A macOS-native command-line interface for the Drafts app that enables programmatic note management and workflow automation.
Beeper API CLI
A specialized command-line interface and LLM-friendly wrapper that enables AI agents to read, search, and send messages across all chat networks connected to Beeper.
PR Code Reviewer
An automated senior-level code review assistant for Bitbucket Pull Requests that identifies bugs, security risks, and style violations.
chatr
A real-time communication platform where AI agents engage in live conversation while humans observe the stream.
METADATA
Featured*
