Passo enables secure remote browser access on your server to handle manual logins, 2FA, and captchas within Openclaw Skills environments.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install passo
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 passo 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 Passo - Remote Browser Access?
Passo is a powerful utility designed to facilitate human-in-the-loop interactions for server-side agents. When integrated as part of your Openclaw Skills, it provides a secure tunnel to a browser instance running on your infrastructure. This allows users to jump into a remote session to perform tasks that are traditionally difficult for AI, such as solving complex captchas or navigating sensitive authentication flows.
By incorporating Passo into your Openclaw Skills setup, you ensure that your automated workflows remain resilient. Instead of failing when encountering a security wall, the agent can provide a secure URL for human intervention, bridging the gap between automated logic and manual verification.
Passo - Remote Browser Access Use Cases
- Managing multi-factor authentication (MFA) or 2FA during automated login sequences.
- Solving interactive captchas that block automated Openclaw Skills agents.
- Logging into web services that require manual Google or social OAuth authentication.
- Real-time debugging of headless browser state on remote production servers.
- Providing a secure manual override for sensitive browser-based tasks.
How Passo - Remote Browser Access Works
- The Passo client is installed on the server where the browser automation is required.
- Use the CLI to start a secure browser tunnel, which initializes Chromium and a VNC instance.
- A unique access URL is generated and shared with the authorized user.
- The user authenticates via Google to gain secure access to the remote browser session.
- After the manual action (like a login) is completed, the agent resumes its automated task using the active session state.
- The session is closed via CLI commands once the Openclaw Skills workflow is finished.
Passo - Remote Browser Access Setup
To add Passo to your Openclaw Skills toolkit, execute the following command on your server:
curl -fsSL https://raw.githubusercontent.com/felipegoulu/passo-client/main/install.sh | bash
Once installed, you can control the browser session using these commands:
passo start # Initiates the remote browser tunnel
passo stop # Terminates all active sessions
passo status # Checks the current connectivity status
Passo - Remote Browser Access Data Schema & Taxonomy
Passo organizes session metadata and security parameters to ensure safe access within Openclaw Skills. The following data points are managed:
| Attribute | Function |
|---|---|
| ACCESS_URL | The dynamically generated link for remote human access. |
| The specific Google account authorized to access the session. | |
| Tunnel Status | Real-time monitoring of the VNC and Chromium processes. |
| Session Persistence | Logic for maintaining browser cookies and state during the hand-off. |
Passo - Remote Browser Access Advanced Features
- Google OAuth protection ensures that only the intended user can access the remote browser.
- Lightweight installation script that automatically handles dependencies like Chromium and VNC.
- Seamless integration with existing Openclaw Skills to provide a robust human-in-the-loop fallback.
- Minimal resource footprint designed for high-performance server environments.
- Automated session management to prevent unauthorized access after the task is completed.
SKILL.md
Loading
Related Openclaw Skills
gcloud
A professional automation skill that enables AI agents to manage Google Cloud Platform resources safely using the official gcloud CLI.
Firestore
A powerful tool for managing Google Cloud Firestore databases directly through the REST API using curl and gcloud authentication.
YouTube Transcript Extraction Skill
A dual-method extraction skill that retrieves YouTube transcripts via CLI tools or automated browser interaction.
xfetch
A cookie-based command-line interface for scraping and retrieving comprehensive data from X/Twitter without requiring API keys.
Ant Design Skill
A comprehensive front-end design skill for building professional React interfaces using Ant Design components, layouts, and v5 design tokens.
Faster Whisper GPU
A high-performance, privacy-first tool for local speech-to-text transcription using NVIDIA GPU acceleration.
Featured*
