#1616th of 24 in Agent frameworks and SDKs
page-agent
In-page JavaScript agent that controls a web page's DOM from natural language
Live demo ↗
(opens in a new tab)Documentation ↗
(opens in a new tab)Repository on GitHub ↗
(opens in a new tab)
- Stars
- 29.4k
- License
- MIT
- Last commit
- Sep 2026
- Last release
- Sep 2026
- Language
- TypeScript
Missed the latest GitHub check; these facts may be older.
Overview
Page Agent is a TypeScript library that embeds an LLM-driven agent in a web page via a script tag or the npm package. It reads the DOM as text, with no screenshots or multimodal models, and executes instructions such as clicking or filling forms. An optional Chrome extension handles multi-page tasks, and a beta MCP server allows external control.
Who it is for: Web developers adding an AI copilot or form automation to their own site
Strengths
- Runs entirely in-page; no Python, headless browser or extension needed for single-page use
- Text-based DOM control, so no vision model or screenshot permissions needed
- Bring your own LLM via baseURL and apiKey config, including locally deployed models
- Installs from npm or a single CDN script
Weaknesses
- Multi-page tasks require the optional Chrome extension
- MCP server is marked Beta
- Demo CDN script uses a free testing LLM API, for technical evaluation only
- API key is passed in client-side config; README gives no guidance on securing it
What it needs
- no GPU
- Needs LLM API endpoint (OpenAI-compatible), zod (peer dependency)
- Models: qwen3.5-plus, OpenAI-compatible APIs, locally deployed models
Also in Agent frameworks and SDKs
See all 24| Rank | Project | Score |
|---|---|---|
| 1 | ppt-masterAgent skill that turns documents into natively editable PowerPoint decks | 91 out of 100 |
| 2 | langchainPython framework for building LLM agents and apps from interoperable components | 88 out of 100 |
| 3 | browser-usePython library that lets an LLM agent control a web browser | 88 out of 100 |
| #4 | tradingagentsLangGraph multi-agent framework simulating a trading firm to analyze tickers | 87 out of 100 |
| #5 | crewaiPython framework for multi-agent workflows with role-based Crews and event-driven Flows | 81 out of 100 |