kimi-webbridge

Automate authenticated browser navigation and interaction via the Kimi WebBridge daemon.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/Z5Jonathan-maker/ai-coding-os --skill kimi-webbridge
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: kimi-webbridge
Source: https://github.com/Z5Jonathan-maker/ai-coding-os/tree/main/claude/skills/kimi-webbridge
Command: npx skills add https://github.com/Z5Jonathan-maker/ai-coding-os --skill kimi-webbridge

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) and scripts (resource) components.

What problem does it solve?

Kimi WebBridge removes the friction of manually operating websites by letting an AI control the user's real browser session to navigate, click, fill forms, read content, and capture screenshots or PDFs.

Core Features & Use Cases

  • Real-session browser automation: Operate any website while using the user's existing login sessions, ideal for tasks that require authenticated access.
  • Accessible page reading & interaction: Use accessibility-tree snapshots to find elements reliably, then click or fill them without brittle CSS selectors.
  • Evidence artifacts for proof: Capture screenshots (via the helper script) and save a rendered page as a PDF for downstream review and record-keeping.
  • Operational resilience: Includes a required health check flow and a dedicated troubleshooting reference to handle missing/disabled daemon or disconnected extension states.
  • Example use case: Have the AI open your web dashboard, navigate to a specific page, fill a form field, submit, and produce a screenshot/PDF as proof of the resulting state.

Quick Start

Ask the AI to take action in the real browser by first running a health check and then navigating to the URL you specify with a new tab in a dedicated session.

Frequently Asked Questions about kimi-webbridge

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate web interaction using my real authenticated browser sessions?

Browser automation with real authenticated sessions is achieved by routing AI commands through the Kimi WebBridge daemon, allowing you to navigate, click, fill forms, and read pages using your existing logins.

How do I extract web content without relying on brittle CSS selectors?

You can extract web content reliably by utilizing accessibility-tree snapshots to find elements on the page, which enables interaction and data extraction without maintaining fragile CSS selectors.

Can I capture screenshots and save PDFs during automated website navigation?

Yes, you can capture screenshots and save rendered pages as PDFs using provided helper scripts, generating evidence artifacts for downstream review and record-keeping during your browser automation tasks.

What should I do if my browser automation task is not working or the daemon disconnects?

If your browser automation task is not working, you should run the required health check flow on the daemon endpoint and consult the dedicated troubleshooting reference to resolve missing daemon or disconnected extension states.

Does web scraping with authenticated sessions require direct base64 screenshot handling?

Web scraping with authenticated sessions does not require direct base64 screenshot handling, as the system provides specific primitives and helper scripts to manage screenshots, navigation, and session lifecycle automatically.