Practical overview
When it fits
Use Opencli when a browser agent needs a command-oriented interface for live websites, especially where an existing Chrome session, structured page interaction or reusable site adapters are useful.
Who it is for
- Developers building AI agents that operate websites.
- Teams creating repeatable browser or website workflows.
- Engineers exposing Electron applications or local tools to agent commands.
- Users who need browser automation through an existing logged-in Chrome session.
Documented uses
- Navigate and inspect live web pages through an AI agent.
- Fill forms, click controls and handle page transitions.
- Extract page data or intercepted network responses.
- Run built-in site commands through a CLI.
- Create reusable website adapters or connect local and desktop tools to agents.
Requirements & limits
The repository documents a desktop app route for macOS and Windows and an npm installation route for CLI-only, CI and server use. npm installation requires Node.js 20.18.1 or newer. A Chrome or Chromium Browser Bridge extension is required for the documented browser-session workflow.
- Browser tasks use a Chrome or Chromium extension and local daemon rather than only a standalone web API.
- The project exposes both built-in adapters and ad-hoc browser primitives.
- A logged-in Chrome session can be used as the browser context; this is marked Verified in the project adaptation notes.
- Creating task-specific website commands is marked Likely compatible because the project describes website-to-CLI conversion without showing a concrete custom implementation in the cited material.
- The documented stack includes JavaScript, Node.js and Playwright.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Set up Opencli with a connected Chrome profile
Install the CLI and Browser Bridge, then verify the browser connection.
Before you start
- Node.js version 20.18.1 or newer for npm installation.
- Chrome or Chromium with the OpenCLI Browser Bridge extension.
- A local installation of Opencli.
Steps
Install Opencli globally with npm using the documented package.
Install the OpenCLI extension from the Chrome Web Store, or load the downloaded release archive as an unpacked extension.
Run
opencli doctorto verify the setup.If multiple Chrome profiles are connected, use the documented profile commands to list, rename and select a profile.
Expected outcome
The local Opencli command is installed and associated with a connected browser profile.
Guide limitations
- The cited setup instructions document installation and diagnosis, not a successful task execution test.
- Profile selection is relevant when multiple connected Chrome profiles are present.
Install Opencli skills for an AI agent
Add the documented Opencli skills so an AI agent can use browser or adapter workflows.
Before you start
- An AI agent environment such as Claude Code or Cursor.
- The
npxcommand available in the development environment. - An Opencli installation and connected browser for browser-driven tasks.
Steps
Run the command below to install the skills.
npx skills add jackwener/opencliAlternatively, install a named skill such as
opencli-browser,opencli-adapter-author,opencli-autofixoropencli-usage.Use the browser skill for navigation, form filling, clicking and extraction, or the adapter-author skill for reusable site commands.
Expected outcome
The selected Opencli skill is installed for the AI agent workflow.
Guide limitations
- The cited material describes skill installation and intended use but does not document an execution result for a particular agent.
- Authenticated workflows depend on the connected logged-in Chrome session described by the project.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 2 October 2026
- Original website title
- jackwener/OpenCLI · Official homepage title · retained in its source language