可交互、可探索的浏览器命令行工具，专为 AI Agent 设计
Project description
NavCLI - Goals & Vision
Core Goal
Enable AI Agents to browse the web like humans.
Existing solutions (HTTP APIs, headless browser scripts, Playwright MCP) have limitations:
- No support for JS-rendered SPAs
- No session persistence
- Lack of interactive exploration
NavCLI's positioning: An interactive, explorable browser CLI
Core Value
| Feature | What NavCLI Solves |
|---|---|
| JS Rendering | Full SPA support |
| Session Persistence | Cookies, session maintained |
| Interactive CLI | Agent can explore while operating |
| Token Optimization | Lightweight elements + on-demand text/html |
Typical Workflow
> g https://example.com # Navigate
> elements # Observe interactive elements
> c .btn-login # Click login
> t #email "test@example.com" # Type email
> t #password "123456" # Type password
> c button[type="submit"] # Submit
> text # Confirm result
Agent can: Navigate → Observe → Interact → Feedback → Continue
Vision
Become the standard browser interaction layer for AI Agents, enabling any Agent to control browsers via command-line interface:
- Form filling, login authentication
- Information scraping, content exploration
- Complex multi-step business processes
Related Documentation
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distributions
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
