[MCP browser-automation] Increase timeout limit from 60000ms

Resolved 💬 1 comment Opened Jan 10, 2026 by trungnguyenvn Closed Jan 10, 2026

Problem

The browser_wait_for_selector tool enforces a maximum timeout of 60000ms (60 seconds). This is insufficient for many real-world browser automation tasks.

Use Cases Requiring Longer Timeouts

| Task | Typical Duration |
|------|-----------------|
| Gemini Deep Research | 2-10 minutes |
| Gemini Image Generation | 30-120 seconds |
| Gemini Video Generation | 60-180 seconds |
| NotebookLM Audio Overview | 60-180 seconds |
| NotebookLM Deep Research | ~5 minutes |

Current Error

MCP error -32000: Tool failed: timeout must be at most 60000ms

Requested Change

Either:

  1. Increase the timeout limit to 600000ms (10 minutes)
  2. Make the limit configurable per-server
  3. Add a browser_wait_for_selector_long variant for extended waits

Workaround

Currently requires implementing polling loops in agent code, which is less efficient and more error-prone than a single wait call.

Related

View original on GitHub ↗

This issue has 1 comment on GitHub. Read the full discussion on GitHub ↗