web_navigate · wangzan101, tools — Claude Code prompt: what it does, how to install
web_navigate — a Claude Code prompt for opening HTTP and HTTPS addresses in Mika's built-in browser. Initiates web navigation before screen tools take over.
Apache-2.0 ★ 9,488 updated 2026-11-16
.claude/commands/web_navigate.mdWhat web_navigate does
The prompt loads web pages strictly through HTTP or HTTPS inside the built-in browser of Mika. Once loaded, screen tools such as web_observe, web_tap, web_type_text, web_scroll, and web_back can interact with the page.
The navigation call waits up to 20 seconds. If the page is still loading, it returns loaded=false, signaling that web_observe should run again. A completed request returns tabId, the final URL, page title, and httpStatus.
When a device rejects plain HTTP, the command automatically retries once over HTTPS. If navigation fails, it yields ok:false with a browser error code such as net::ERR_NAME_NOT_RESOLVED. Target URLs pointing to files trigger an automatic download instead, returning the download status.
What it helps with
- open web addresses inside Mika's built-in browser
- retrieve
tabIdand page load status - download files when navigating to file URLs
- handle connection failures and plain HTTP retries over HTTPS
Example prompts for the agent
- Open the project documentation in the Mika browser
- Navigate to the link from the task and wait for the page to load
How to install web_navigate
Manually: put the files in place
.claude/commands/web_navigate.mdHow to use
Paste the text at the start of a chat or into project instructions
Questions about web_navigate
How to install web_navigate in Claude Code?
Place the prompt file at .claude/commands/web_navigate.md in your project.
What does web_navigate return?
It returns tabId, the final URL, title, and httpStatus, or ok:false alongside a netError code on failure.
How does web_navigate handle timeouts and downloads?
If loading exceeds 20 seconds, it returns loaded=false for web_observe to check again. Direct file URLs trigger downloads with status reported back.
Source text
web_navigate.md · Apache-2.0
Open a web page (http or https only) in Mika's built-in browser; the web_* screen tools (web_observe, web_tap, web_type_text, web_scroll, web_back, ...) then act on it. Waits up to 20 seconds for the page to load (loaded=false: still loading, web_observe again). Returns tabId, the final url, title and httpStatus. If the page cannot be opened, it returns ok:false with the browser's error code (netError, e.g. net::ERR_NAME_NOT_RESOLVED); an http address the phone refuses as plain text is tried once over https. If the address turns out to be a file, it is downloaded instead and the result says so with the download's status.
© wangzan101 · Apache-2.0 · wangzan101/Mika
Similar prompts
All prompts for Browser and scraping →web_navigate by other authors comes first.