Tactical

Daily Telegram status for a Python bot

You should not log into the exchange to see if the process is still alive. A dull 08:00 ping is enough. A green percent is not the product.

This is a health ping. It is not a P&L dashboard and not financial advice. Paper-trade has no real profit. Trading can lose money.

What the message should look like

openclaw  08:00 UTC
dry_run=true
halt=absent
loss 0.00 / 10.00
last scan 07:55  71 markets
open orders: 0
unit=active
errors since midnight: 0

If dry_run flipped, halt appeared, unit is dead, or open orders do not match what you think you have — that is the ping that matters. Not “Daily PnL: +2.5%.”

What to collect

• systemctl is-active openclaw (or equivalent).
• DRY_RUN from the same env the service uses.
• HALT file exists or not.
• Daily loss vs limit (your kill-switch counters).
• Last successful scan line from the log or an in-process timestamp.
• Count of open orders on the venue (empty + a position = problem).

Optional later: fill count and fees if you are live. Put that last. Never as the title or the thumbnail.

Send the ping

import os, json, urllib.request, urllib.parse

def telegram(text: str) -> None:
    token = os.environ["TELEGRAM_BOT_TOKEN"]
    chat = os.environ["TELEGRAM_CHAT_ID"]
    url = f"https://api.telegram.org/bot{token}/sendMessage"
    body = urllib.parse.urlencode({"chat_id": chat, "text": text}).encode()
    urllib.request.urlopen(url, data=body, timeout=15)

Same token and chat as your fill/fail alerts. If this curl fails, fix Telegram first — do not add a timer on a silent bot.

Trigger: systemd timer, not while True

Do not start a second infinite loop. The bot already has a scan loop. A timer unit runs the status script at 08:00 UTC and exits.

# /etc/systemd/system/openclaw-status.service
[Unit]
Description=OpenClaw morning status
After=network-online.target

[Service]
Type=oneshot
User=YOURUSER
WorkingDirectory=/home/YOURUSER/bots/openclaw
EnvironmentFile=/home/YOURUSER/bots/openclaw/.env
ExecStart=/home/YOURUSER/bots/openclaw/.venv/bin/python status.py

# /etc/systemd/system/openclaw-status.timer
[Unit]
Description=OpenClaw morning status timer

[Timer]
OnCalendar=*-*-* 08:00:00
Persistent=true
Unit=openclaw-status.service

[Install]
WantedBy=timers.target

sudo systemctl daemon-reload
sudo systemctl enable --now openclaw-status.timer
systemctl list-timers | grep openclaw

Optional: /status on demand

If you already poll getUpdates, reply to /status with the same payload as 08:00. Do not add a webhook and a second process until the timer ping is boring for a week.

If Telegram is silent, fix that first: /guides/telegram-not-sending. systemd itself: /guides/systemd-journalctl.

Keep reading