A domain-restricted local search engine that provides authoritative technical documentation results without web spam.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install curated-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 curated-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 Curated Search?
Curated Search is a specialized tool designed to provide developers with a clean, authoritative search experience over a whitelist of technical documentation, such as MDN, Python docs, or specialized man pages. By focusing on specific domains, this Openclaw Skills integration eliminates the noise and SEO-driven spam often found in general search engine results.
The skill operates with a local-first philosophy. Once the initial index is built, all search operations are performed entirely on your machine. This ensures high-speed query resolution and absolute privacy, as no search data or workspace context ever leaves your local environment.
Curated Search Use Cases
- Searching for specific API syntax across MDN or official language documentation without browser distractions.
- Accessing authoritative technical guides while avoiding low-quality forum posts or outdated blog content.
- Performing offline documentation searches in environments with limited or no internet connectivity.
- Filtering search queries to specific trusted domains to ensure language-specific technical relevance.
How Curated Search Works
- The user defines a whitelist of trusted domains and seed URLs in a local configuration file.
- A manual crawl process is initiated to scrape and index content from the approved domains.
- The indexed data is stored locally in a compressed format on the host machine.
- When a query is made, the tool utilizes a BM25 relevance algorithm to scan the local index.
- Structured JSON results, including snippets and relevance scores, are returned to the agent for immediate use.
Curated Search Setup
To get started with this skill for Openclaw Skills, follow these installation steps:
- Ensure you have Node.js installed on your system.
- Configure your target documentation sources by editing the
config.yamlfile in the skill directory. - Build your local search index by running the crawler command:
npm run crawl
- Once the crawl is complete, the skill is ready to handle search queries through the agent interface.
Curated Search Data Schema & Taxonomy
The skill returns a JSON array of result objects. The data is organized according to the following schema:
| Field | Type | Description |
|---|---|---|
title |
string | The cleaned document title from the source. |
url |
string | The canonical source URL of the documentation. |
snippet |
string | A ~200 character excerpt from the content. |
domain |
string | The hostname of the source (e.g., docs.python.org). |
score |
number | The BM25 relevance score for the query match. |
crawled_at |
number | Unix timestamp indicating when the page was last indexed. |
Curated Search Advanced Features
- Domain-restricted filtering to isolate searches to specific documentation sets.
- Adjustable relevance thresholds using the min_score parameter to ensure only high-quality matches appear.
- Full pagination support via offset and limit parameters for deep index exploration.
- Local-only execution path ensuring queries within Openclaw Skills remain private and secure.
- Configurable crawling parameters including depth, delays, and content length limits to customize index density.
SKILL.md
Loading
Related Openclaw Skills
ClawShop
An API-driven skill for managing product listings and metadata on Taobao and Goofish marketplaces.
Monkeytype Tracker
A powerful typing performance monitor that integrates Monkeytype statistics and personalized improvement tips into your AI agent workflow.
Anthropic Frontend Design
A high-performance design skill for building production-grade frontend interfaces that reject generic AI patterns in favor of bold, intentional aesthetics.
Wind & Site (Wind Rose & Data)
Fetch real-time wind data and generate professional wind roses for comprehensive site and urban climate assessments.
Session State Tracker
A robust lifecycle management tool that prevents context loss by automatically persisting and restoring AI agent state across compactions and restarts.
Token Optimizer
A comprehensive toolkit designed to drastically reduce token usage and API expenditures through intelligent context management and task-based model routing.
METADATA
Featured*
