Sport Mode

Temporarily boost heartbeat frequency (default 3m) and inject a monitoring task into HEARTBEAT.md. Perfect for supervising background agents (Codex), long-running builds, or interactive games.

Usage

# Turn ON: Set heartbeat to 3m and set monitoring task
sport-mode on --task "Check Codex progress. If done, run sport-mode off."

# Custom Interval: Set to 1 minute
sport-mode on --task "Game tick" --every "1m"

# Turn OFF: Reset heartbeat to 30m and clear HEARTBEAT.md
sport-mode off

How it works

ON:
- Patches ~/.openclaw/openclaw.json (hot-reload) to set heartbeat.every.
- Writes your task to HEARTBEAT.md with a "Sport Mode Active" header.
OFF:
- Patches config back to 30m (default).
- Clears HEARTBEAT.md.

Best Practices

1. Set a Finish Line

Unless you want an endless marathon, always define a termination condition in your task.

✅ Good: "Monitor build. If success or fail, run sport-mode off."
❌ Bad: "Monitor build." (Agent might keep reporting "Done" forever until you manually stop it).

2. State Machine in File

For multi-step tasks (like games or staged deployments), let the agent update HEARTBEAT.md itself.

Pattern: Read state -> Execute step -> Write new state -> Sleep.
This keeps the agent "stateless" (doesn't rely on conversation history context window) but the task "stateful".

3. Use tmux for Visibility

If the monitoring task involves terminal output (e.g., Codex coding, compiling), running the task in a tmux session is ideal.

The agent can inspect the pane (tmux capture-pane) without interfering.
The user can attach (tmux attach) to watch live.

4. Silence is Golden

For high-frequency modes (e.g., 1m), avoid spamming "Nothing happened".

Configure the agent to reply HEARTBEAT_OK (silence) if the status hasn't changed.
Only notify the user on milestone completion, errors, or final success.

Implementation Note

This skill uses openclaw config set to safely patch configuration at runtime, triggering a seamless Gateway reload.

openclaw-sport-mode