Browser tool: URLs with Chinese characters are mis-encoded

Symptoms

• Using the browser tool with a targetUrl that includes Chinese characters in query parameters navigates to the wrong page/query.
• The final URL shows percent-encoding that maps to different characters than what you intended.

Cause

The browser tool incorrectly encodes non-ASCII characters in URL query parameters (not producing the expected UTF-8 percent-encoding).

Fix

Until OpenClaw fixes the upstream encoding bug, use one of these workarounds so the browser tool only receives ASCII URLs.

1) Pre-encode the query parameter yourself (recommended)

Build the URL with proper UTF-8 encoding (and avoid double-encoding).

JavaScript / TypeScript:

const q = "AI革命";
const url = "https://www.google.com/search?q=" + encodeURIComponent(q);
// Pass `url` to the browser tool.

Python:

import urllib.parse

q = "AI革命"
url = "https://www.google.com/search?q=" + urllib.parse.quote(q, safe="")

2) If you already have a correct percent-encoded URL, do not re-encode it

If your URL already contains %E9%9D%A9... style sequences, pass it through as-is.

Verify

• The browser navigates to the intended query/page.
• The URL uses the expected UTF-8 percent-encoding for the Chinese characters (for example, "AI革命" encodes to %E9%9D%A9%E5%91%BD).

Related

• GitHub: #9574