Skip to content

[Feature] Nanobrowser Integration of features #419

@sgtest2026-ux

Description

@sgtest2026-ux

Feature Description / 功能描述

I generally use NanoBrowser to log in to a webpage and then click through all the menus one by one, but there hasn't been any development on NanoBrowser for a long time. When I checked out this project, it's really nice, and I think it would be even better if NanoBrowser features were added to it.

https://github.com/nanobrowser/nanobrowser

Nanobrowser vs Alibaba Page-Agent

Category Nanobrowser Alibaba Page-Agent
Type Chrome Extension Chrome Extension/JavaScript Library / Framework
Core Role External AI agent controlling browser Embedded AI agent inside web app
Target Audience End-users Developers / Product teams
Deployment Browser extension install Script tag / npm integration
Execution Environment Browser (outside page) Inside webpage (client-side)
Scope Works across entire web Limited to single application
Control Model Controls browser like a human user Direct DOM-level control
Architecture Multi-agent system supported :contentReference[oaicite:0]{index=0} DOM-first, text-based interaction :contentReference[oaicite:1]{index=1}
Multi-site Capability Strong (cross-site workflows) Limited (unless extended)
DOM Interaction Yes Yes (direct DOM parsing)
Vision / Screenshot Use Often used in agents Not required (DOM-based) :contentReference[oaicite:2]{index=2}
Automation Type General web automation In-app automation
Integration Works on any website Must be embedded into app
Setup Install extension Add script or npm package :contentReference[oaicite:3]{index=3}
LLM Support Flexible / multi-model BYOLLM (bring your own model) :contentReference[oaicite:4]{index=4}
Use Cases Scraping, browsing, automation, research Copilot, onboarding, form automation
UI Interaction Yes (simulates user actions) Yes (direct UI manipulation)
Speed Depends on agent + browser Faster (no vision layer) :contentReference[oaicite:5]{index=5}
Product Integration ❌ Not designed for embedding ✅ Designed for embedding
Developer Control Low High
Privacy Model Often local-first Controlled via chosen LLM
Open Source Yes Yes :contentReference[oaicite:6]{index=6}
Mental Model AI uses the internet for you AI helps users inside your app
Best For Automation across websites Adding AI into SaaS / products
Not Suitable For Embedding into apps Cross-site automation

One-Line Summary

  • Nanobrowser → External AI agent (automation across the web)
  • Page-Agent → Internal AI agent (copilot inside your app)

Final Verdict

These tools are not direct competitors:

  • Choose Nanobrowser if you need autonomous browsing & automation
  • Choose Page-Agent if you need embedded AI inside a product

Before submitting

  • I will be polite and respectful. / 我会保持礼貌与尊重。
  • I have read the CODE_OF_CONDUCT.md and CONTRIBUTING.md. / 我已阅读行为准则。
  • I have searched existing issues and this is not a duplicate.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions