# alibaba/page-agent

JavaScript in-page GUI agent. Control web interfaces with natural language.

Repository: https://github.com/alibaba/page-agent
Canonical: https://ross.abutalabs.com/products/page-agent
Homepage: https://alibaba.github.io/page-agent/
Language: TypeScript
License: MIT
License Family: permissive
Topics: agent, ai, web, ai-agents, javascript, typescript, browser-automation, mcp
Last push: 2026-08-24T19:09:23+00:00

## Health v2 (maintenance only)
Score: 82/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 93, longevity 24
- inputs: {"age_days": 344, "days_push": 9, "days_rel": 48, "gap_med": 3.0, "n_releases_24m": 31}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 28847, forks 2580 (observed 2026-08-28T04:11:52.906561+00:00)

## What it is
Page Agent is a JavaScript library that embeds an AI GUI agent directly into any web page with a single script, letting users control the interface using natural language. It works via text-based DOM manipulation without screenshots, browser extensions, or multimodal LLMs, and supports bringing your own LLM including locally deployed models.

## Use cases
- add a natural language AI agent to my web app
- automate web page interactions with plain English commands
- let users control my site's UI by chatting
- browser automation without headless browsers or extensions
- embed an in-page GUI agent driven by my own LLM
- script DOM tasks from natural language instructions

## When to choose
- you want in-page AI automation without extensions, Python, or headless browsers
- you want text-based DOM control that avoids multimodal models and screenshot costs
- you need to bring your own LLM, including locally deployed ones
- you want a Chrome extension option for injecting the agent into arbitrary sites

## When to avoid
- you need pixel-level visual automation requiring screenshots or vision models
- you need cross-application desktop automation outside the browser
- your app cannot expose its DOM or run third-party JavaScript
- you require fully deterministic scripted automation rather than LLM-driven actions

## Facets
- artifact type: library
- maturity: active
- function: agent-framework, browser-extension, llm-inference, workflow-automation
- domain: web-development, browser-extensions, large-language-models
- platform: browser, browser-extension
- tags: gui-agent, dom-automation, in-page-agent, natural-language-control, web-automation, mcp, ai-agents, automation, web-server, nodejs, javascript

## Member repositories
- alibaba/page-agent (main) score 82

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:52.906561+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T16:53:17.172219+00:00, confidence not recorded.
  - readme: https://github.com/alibaba/page-agent (fetched 2026-08-28T04:11:52.906561+00:00, sha 6a107c4b4b3a)
  - homepage: https://alibaba.github.io/page-agent/ (fetched 2026-08-29T07:50:53.549174+00:00, sha 3ca30c368ff4)
  - registry_npm: https://registry.npmjs.org/page-agent (fetched 2026-08-29T07:50:53.551868+00:00, sha e07731b06616)
- Data as of 2026-08-30T08:39:29.467469+00:00.