Summary

Manual cron job execution via openclaw cron run <jobId> does not work on macOS. The gateway receives the request and queues it, but never executes — times out after 60 seconds.

Environment

• OpenClaw versions tested: v2026.2.23, v2026.2.26
• Gateway: local loopback 127.0.0.1:18789

Steps to reproduce

1. Create any cron job (isolated or main)
2. Run openclaw cron run <jobId> or trigger from agent via cron(action="run")
3. Wait → timeout after 60s, job never executes

What works ✅

• cron add, list, status, remove, runs
• Scheduled automatic execution (jobs run fine at their scheduled time)

What fails ❌

• cron run <jobId> (both CLI and tool call)
• Tested with --due flag and without
• Tested with both job types: isolated (agentTurn) and main (systemEvent)

Log behavior

1. Gateway receives request and enqueues to lane:
   • lane enqueue: lane=session:agent:main:cron:<jobId>
   • lane dequeue immediate
2. Then nothing — no execution log, no error, no stack trace
3. Job marked as "running" but never starts
4. On gateway restart: cron: clearing stale running marker on startup

From agent

Tool call never receives response → missing tool result in session history

From CLI

gateway timeout after 60000ms

Additional context

• When triggered from agent, the long timeout seems to cause gateway restart (conflict with SIGUSR1?)
• cron runs --id <jobId> shows 0 entries — job never actually ran
• No crashes or stack traces, execution simply never starts after dequeue