OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Browser Agents

Page Agent

A TypeScript/JavaScript library that places a GUI agent inside a web page and controls interfaces through natural-language instructions.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Page Agent — OpenViST project profile in Browser Agents.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Page Agent when the intended automation belongs in the web page itself and can be expressed through text-based DOM interaction. Consider its optional browser extension or beta MCP Server when the workflow needs broader browser control.

Who it is for

  • Developers adding AI controls to existing web applications
  • Teams building in-page assistants for repetitive interface workflows
  • Developers working on browser automation, form filling, or accessibility-oriented interaction

Documented uses

  • Add natural-language controls to a web application
  • Automate repetitive GUI actions such as form and interface workflows
  • Connect page-level agent behavior to a developer-selected language model
  • Extend an agent across browser tabs with the optional Chrome extension
Before you choose

Requirements & limits

The repository is public and licensed under MIT. The project provides an npm package, a demo, and documentation. The demo CDN is identified as a technical-evaluation path using a testing LLM API.

  • Basic integration is in-page JavaScript and does not require a browser extension, Python, or headless browser.
  • The documented interaction model uses text-based DOM manipulation rather than screenshots or multimodal LLMs.
  • Developers can bring mainstream or locally deployed language models.
  • The optional Chrome extension supports multi-page tasks, while the MCP Server is marked beta.
  • The project is intended for client-side web enhancement, not server-side automation.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Add a basic in-page Page Agent

Install Page Agent and create an agent that executes a natural-language web-interface instruction.

Before you start

  • A TypeScript or JavaScript web application
  • npm
  • An LLM model, API base URL, and API key for the package-based example

Steps

  1. Install the package with npm install page-agent.

  2. Import PageAgent from page-agent.

  3. Create a PageAgent instance with a model, baseURL, apiKey, and language.

  4. Call execute with a natural-language instruction such as Click the login button.

Expected outcome

The application has a configured Page Agent instance that can receive a natural-language GUI instruction.

Guide limitations

  • The documented demo CDN uses a free testing LLM API and is for technical evaluation only.
  • The example documents client-side web enhancement, not server-side automation.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
1 October 2026
Original website title
alibaba/page-agent · Official homepage title · retained in its source language
How we review sourcesSuggest a correction