Skip to content

gisk0/claude-watchdog

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

40 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Claude Watchdog 🐕

Know when Claude goes down — before it ruins your workflow.

An OpenClaw skill that monitors the Claude API for outages and latency spikes, sending rich alerts straight to Telegram.

What You Get

Status alerts when Anthropic reports incidents:

🟠 Anthropic Status: Partially Degraded Service

📌 Elevated error rates on Claude 3.5 Sonnet (⚠️ affects us)
Status: Investigating
Update: "We are investigating increased error rates..."

Components:
  🟠 API: partial outage

🔗 https://status.claude.com

Latency spike detection with severity levels:


🟡 Anthropic API — High Latency Detected

Current: 12.3s
Baseline: 3.1s (median of last 19 samples)
Ratio: 4.0×

Recovery notifications so you know when it's safe to go back:

✅ Anthropic API — Latency Back to Normal


Current: 2.8s
Baseline: 3.1s
Was: 12.3s when alert fired

Features

  • Status monitoring — polls status.claude.com every 15 min with incident details, affected components, and model relevance tagging ("affects us" vs "not our model")
  • Latency probing — sends a minimal request through your OpenClaw gateway, maintains a rolling baseline, alerts on spikes with 🟡/🟠/🔴 severity
  • Recovery alerts — notifies when incidents resolve and latency returns to normal
  • Telegram topic support — target a specific forum topic for alerts
  • Zero dependencies — stdlib Python only, no pip install needed
  • Tunable thresholds — adjust alert sensitivity, baseline window, and probe timeout
  • Secure by default — all config files chmod 600, secrets never logged

Prerequisites

  • OpenClaw running on your server
  • Python 3.10+
  • A Telegram bot (free — create one via @BotFather)

Install

clawhub install claude-watchdog

The interactive setup walks you through connecting your Telegram bot and OpenClaw gateway. It sends a test alert at the end so you know it works.

Cost

Component Cost
Status checks $0 — scrapes public status page
Latency probes ~$0.000001/probe — minimal "Reply OK" request
Total/month < $0.01

How It Works

┌─────────────┐    every 15 min    ┌──────────────────────┐
│  cron job    │──────────────────▶│  status-check.py     │
│              │                   │  • polls status.claude│
│              │                   │  • diffs vs last state│
│              │                   │  • alerts on change   │
└─────────────┘                   └──────────────────────┘
                                          │
                                          ▼ Telegram API
                                    ┌──────────┐
                                    │  Your TG  │
                                    │  chat/topic│
                                    └──────────┘
┌─────────────┐    every 15 min    ┌──────────────────────┐
│  cron job    │──────────────────▶│  latency-probe.py    │
│              │                   │  • pings OpenClaw GW  │
│              │                   │  • rolling baseline   │
│              │                   │  • alerts on spike    │
└─────────────┘                   └──────────────────────┘

No tokens consumed for status checks. Latency probes cost fractions of a cent per month.

Docs

See SKILL.md for the full configuration reference, alert tuning guide, and state file details.

License

MIT

About

OpenClaw skill: Anthropic API status monitor + latency probe with Telegram alerts

Resources

License

Stars

Watchers

Forks

Packages

 
 
 

Contributors