Bug type

Behavior bug (incorrect output/state without crash)

Beta release blocker

No

Summary

Telegram polling startup emits the normal [telegram][diag] polling cycle started ... diagnostic through the error output path, making normal startup look like a gateway error.

Steps to reproduce

1. Start the Telegram plugin in polling mode.
2. Observe the startup diagnostic line for a polling cycle.
3. Confirm the normal diagnostic is rendered through the error channel.

Expected behavior

Normal polling lifecycle diagnostics should be emitted through the normal log channel, not the error channel. Only actual Telegram monitor failures should use the error channel.

Actual behavior

The observed startup diagnostic is a normal lifecycle line but is rendered with error styling in console output.

OpenClaw version

NOT_ENOUGH_INFO

Operating system

NOT_ENOUGH_INFO

Install method

NOT_ENOUGH_INFO

Model

NOT_ENOUGH_INFO

Provider / routing chain

NOT_ENOUGH_INFO

Additional provider/model setup details

NOT_ENOUGH_INFO

Logs, screenshots, and evidence

Trace/proof:
- gateway-dev.log line 56:
  "[telegram] [diag] polling cycle started inFlight=0 outcome=not-started startedAt=n/a finishedAt=n/a durationMs=n/a offset=n/a"
  rendered with ERROR coloring in console output.

Impact and severity

Affected: Telegram polling users/operators watching gateway logs.
Severity: Medium; normal startup looks like a gateway error and can trigger false alarms.
Frequency: Observed 4 matching startup diagnostic lines; at least one was rendered as an error.
Consequence: Operators can misread healthy Telegram polling startup as a failure.

Additional information

Current code path: extensions/telegram/src/monitor.ts wired the polling monitor logger to runtime.error/console.error, while extensions/telegram/src/polling-session.ts sends the normal polling cycle started lifecycle diagnostic through that logger.