A secure, CLI-driven browser automation skill that enables AI agents to search, crawl, and fetch web page data via the Hyperbrowser platform without managing API keys directly.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install oo-hyperbrowser
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 oo-hyperbrowser 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 Hyperbrowser?
The oo-hyperbrowser skill allows developers and AI agents to seamlessly interact with Hyperbrowser (hyperbrowser.ai) using OOMOL-connected accounts. By utilizing this skill, you bypass the complexity of manually managing raw API keys and tokens. The credentials are secure and injected server-side by OOMOL, creating a highly robust environment for web-scraping pipelines.
As a core addition to your collection of Openclaw Skills, this tool acts as a unified hub for reading, creating, and updating web automation data. Whether you need to scrape dynamic websites, execute deep web crawls, or perform structured search queries, this skill provides a clean CLI-based interface for your AI workflows.
Hyperbrowser Use Cases
- Automating massive web-scraping pipelines by initiating and tracking asynchronous crawl jobs.
- Extracting raw structured text or media from specific target URLs using headless browser technology.
- Performing real-time, structured web searches to feed live data directly into LLM agents.
- Running safe read-only operations alongside sensitive write operations that require guardrails and explicit approvals.
How Hyperbrowser Works
- CLI Execution: The skill leverages the OOMOL
ooCLI tool under the hood, utilizing safe bash-based execution workflows. - Schema Inspection: The AI agent inspects the authoritative data schema for a given action using
oo connector schemato ensure payload accuracy. - Server-Side Token Injection: OOMOL automatically handles and injects the necessary Hyperbrowser credentials without exposing them in scripts or local environment variables.
- Execution: The action is run with the prepared JSON payload, and the results are returned as a standard JSON object containing the execution ID and raw action output.
Hyperbrowser Setup
To begin using this capability within your Openclaw Skills workspace, follow these essential setup steps:
1. Install the oo CLI
If the command oo is not found, install it using the following commands:
macOS / Linux:
curl -fsSL https://cli.oomol.com/install.sh | bash
Windows PowerShell:
irm https://cli.oomol.com/install.ps1 | iex
2. Sign In to OOMOL
Authenticate your local environment by running:
oo auth login
3. Connect Hyperbrowser
Link your Hyperbrowser account at the OOMOL App Connections page:
https://console.oomol.com/app-connections?provider=hyperbrowser
(Auth Type: API Key)
Hyperbrowser Data Schema & Taxonomy
This skill exposes multiple core actions via the oo command-line execution framework. You can inspect the structure of each action prior to running it.
Schema Inspection Command
oo connector schema "hyperbrowser" --action "<action_name>"
Available Actions & Schema Overview
| Action Name | Type | Description | Safe to Run Directly? |
|---|---|---|---|
fetch_page |
Read | Fetches a target web page and returns it in requested data formats. | Yes |
get_web_crawl_results |
Read | Retrieves paginated results for a completed or ongoing crawl job. | Yes |
get_web_crawl_status |
Read | Checks the current status (pending, processing, completed) of a crawl. | Yes |
search_web |
Read | Executes structured search engine queries and returns results. | Yes |
start_web_crawl |
Write | Initiates an asynchronous crawl job starting from a root URL. | No (Requires User Consent) |
Response Payload Structure
All connector executions return JSON matching the following structure:
{
"data": { ... },
"meta": {
"executionId": "string"
}
}
Hyperbrowser Advanced Features
- Server-Side Credential Protection: Completely eliminates the risk of leaking private API keys in client environments.
- Write-Action Guardrails: Distinguishes between safe read operations and state-altering write operations (
start_web_crawl), prompting explicit human confirmation before execution. - Dynamic Schema Inspection: Dynamically fetches authoritative JSON schemas for all endpoints, ensuring compatibility even if upstream payloads update.
- Error Boundary Controls: Integrates clear billing and authentication troubleshooting alerts, letting your agent automatically recover from errors like token expiration or credit depletion.
SKILL.md
Loading
Related Openclaw Skills
oo-neverbounce
An executive API bridge for NeverBounce to run high-accuracy email validation and bulk verification tasks directly through the OOMOL CLI.
oo-loomio
A developer-ready CLI-based skill that allows AI agents to interact directly with Loomio to search, list, and retrieve poll details safely.
lexoffice
A powerful integration to operate and automate lexoffice accounting, contact, and inventory workflows via OOMOL secure CLI actions.
Klangio AI Audio Transcription and Analysis
An AI-powered audio transcription and analysis skill that automates beat tracking, chord recognition, source separation, and guitar strum detection.
Fidel API
A secure Openclaw Skills connector for query, transaction, and card data retrieval via the Fidel API using the OOMOL CLI tool.
Formbricks Integration
An AI-ready integration to programmatically manage Formbricks contacts, attribute keys, and workspaces using the oo CLI without handling raw tokens.
Featured*
