Bug type

Regression (worked before, now fails)

Beta release blocker

No

Summary

The "Communications", "AI & Agents", and "Automation" menus in the OpenClaw dashboard-v2 are extremely slow, often freezing the UI for 10-15 seconds. This is directly related to Issue #45698 and is caused by blocking health probes (getMe) for Telegram and heavy config schema rendering on the frontend.

Steps to reproduce

1. Open OpenClaw dashboard-v2.
2. 2. Click on the "Communications", "AI & Agents", or "Automation" menu.
3. 3. Observe the browser UI freezing for 10-15 seconds before the page content finally appears.
4. 4. Verify in gateway logs that the Telegram health probe (getMe) timed out due to an IPv6 blackhole or high latency.

Expected behavior

The dashboard menus should load asynchronously and remain responsive. A delay or timeout in a specific backend channel (e.g., Telegram API) should not block the entire gateway's event loop or freeze the user interface.

Actual behavior

A blocking health probe (getMe) for Telegram triggers a 10s timeout, stalling the entire gateway's main event loop. Concurrent WebSocket requests (models.list, sessions.list, etc.) are forced to wait. After the data is finally returned, the dashboard-v2 frontend freezes for another 3-5s while attempting to render a 120KB configuration payload combined with a large schema.

OpenClaw version

2026.5.2

Operating system

Linux (Docker)

Install method

docker

Model

any (e.g. minimax/text-01)

Provider / routing chain

openclaw -> telegram (blocking health probe)

Additional provider/model setup details

The issue is specifically triggered by the Telegram channel health monitor. When 'healthMonitor.enabled' is true, the 'getMe' probe blocks the gateway's event loop for 10 seconds (hardcoded timeout). This stalls all other concurrent WebSocket requests.

Architecture Recommendations:

1. Make health probes asynchronous/non-blocking so they don't stall the main event loop.
2. 2. Expose 'httpTimeoutMs' for Telegram channels (similar to the Feishu implementation).
3. 3. Optimize dashboard-v2 form rendering for large configuration schemas to prevent UI freezing.

Logs, screenshots, and evidence

Impact and severity

Affected: All users using Telegram channel with health monitoring enabled.
Severity: High (Makes dashboard unusable).
Frequency: Always when accessing specific menus (Communications, AI & Agents, Automation).
Consequence: 10-15s UI freeze and gateway-wide main loop blocking.

Additional information

Workarounds identified:

1. Set channels.telegram.healthMonitor.enabled = false to disable the blocking getMe probe.
2. 2. Use NODE_OPTIONS=--dns-result-order=ipv4first to bypass IPv6 blackhole issues that trigger the 10s timeout.
      Architecture Recommendations:
3. Decouple health probes from the main request flow (make them non-blocking).
4. 2. Expose httpTimeoutMs for Telegram channels (similar to Feishu) to allow user-defined timeouts.
5. 3. Optimize dashboard-v2 frontend rendering for large config schemas.