⚡ Blip

Free, serverless uptime monitoring & status pages — powered by Cloudflare Workers.

$0/month. One Worker. No server to run.

Quick start · Docs · Configuration · Deploy · FAQ

---

Blip is a single Cloudflare Worker that monitors your websites, APIs, TLS certificates, and domains on a cron trigger, stores results in Cloudflare D1 (managed SQLite), and serves a beautiful React status page + admin dashboard straight from the same Worker. Built for agencies, teams, and homelabbers: multi-tenant groups, RBAC, and client-scoped dashboards out of the box.

✨ Features

• 🆓 Truly $0/month — one Worker + D1 on Cloudflare's free tier. No server, no VM, no cron host, no database to run.
• 🌍 External monitoring every 5 minutes from Cloudflare's edge network.
• 🏠 Monitor your homelab with no open ports — reach LAN-only services through a Cloudflare Tunnel + Access service token. Setup →
• 🔎 Rich checks — HTTP status, keyword present/absent, JSON-body assertions, raw TCP ports, TLS cert expiry, and domain expiry.
• ⚡ Fast dashboard — React + Vite SPA with 24h / 7d / 30d / 90d graphs and incident history, served from the edge.
• 🏢 Multi-tenant — group sites by client/product/region.
• 🔐 Built-in RBAC — SUPER_ADMIN / ADMIN / CLIENT / VIEWER, password auth at the edge, with client-scoped views.
• 🔔 Notifications — Telegram, Email (Resend), Discord, Slack, and a generic webhook, with per-channel routing + flapping suppression.
• 🎨 White-label ready — your name, colors, logo on a custom domain.
• 🧩 One typed config — a single blip.config.yaml, validated against shared TypeScript types.

🖼️ Screenshots

Public status page

A branded, standalone page you can share with users — grouped components and 90-day uptime bars.

Admin dashboard

Overview	Site detail
Sites	Incidents

🏗️ Architecture

flowchart TD
    CFG[blip.config.yaml] -->|embedded at build| WORKER

    subgraph WORKER["Cloudflare Worker"]
        CRON["scheduled() · cron */5<br/>probe sites: HTTP · TCP · SSL · domain"]
        FETCH["fetch() · serves dashboard SPA<br/>+ /data/*.json API"]
    end

    CRON -->|append points + blobs| D1[("Cloudflare D1<br/>summary · history · incidents")]
    CRON -->|alerts| NOTIFY["Telegram · Email · Discord<br/>Slack · webhook"]
    D1 --> FETCH
    FETCH --> USERS["Users · status page · admin dashboard"]

Loading

ASCII fallback

blip.config.yaml  (embedded into the Worker at build time)
      │
      ▼
┌──────────────── Cloudflare Worker ────────────────┐
│ scheduled() (every 5 min) ── probe sites          │
│      │                  │                          │
│  write to D1       send alerts → Telegram/Email/…  │
│      ▼                                             │
│ fetch() ── serves dashboard SPA + /data/*.json API │
└────────────────────────────────────────────────────┘
      ▼
Cloudflare D1 (managed SQLite) ── summary · history/<id> · incidents
      ▼
Users · status page · admin dashboard  (RBAC + password auth at the edge)

Full write-up: docs/architecture.md.

🚀 Quick start

# 1. Use this template → create your repo → clone it, then:
npm install

# 2. Generate your config interactively (brand, sites, channels)
npm run setup

# 3. Preview locally with realistic demo data
npm run seed && npm run dev      # → http://localhost:5173

# 4. Deploy to Cloudflare — one command does it all
npx wrangler login               # once
npm run deploy:cloud             # creates D1, sets secrets, deploys 🎉

deploy:cloud creates the D1 database and writes its id into wrangler.toml, applies the schema, asks for your domain, generates the session key, prompts for your admin password + any channel tokens, and deploys the Worker. The Worker then probes every site on its 5-minute cron and serves the dashboard + status page from the same URL.

Handy commands: npm run build (build dashboard) · npm run typecheck · npm test.

⚙️ Configuration

Everything lives in one file, blip.config.yaml:

version: 1
brand:
  name: Blip
  primaryColor: "#22c55e"
sites:
  - name: My Site
    url: https://example.com
    ssl: true            # watch TLS cert expiry
    domain: true         # watch domain expiry
  - name: My API
    url: https://api.example.com/health
    expectedStatus: 200
    keyword: "ok"        # body must contain "ok"
    expectJson:
      - path: "db.connected"
        equals: true
channels:
  - id: telegram-main
    type: telegram
    botToken: ${TELEGRAM_BOT_TOKEN}   # wrangler secret put TELEGRAM_BOT_TOKEN
    chatId: ${TELEGRAM_CHAT_ID}
    events: [down, up, ssl, domain]

Full reference (every field, every check type, every channel): docs/configuration.md. A complete annotated example lives at config/blip.config.example.yaml.

🔔 Notifications

Five channels, each with events / sites / groups filters and minDownMinutes flap suppression:

Channel	Secret(s)
Telegram	TELEGRAM_BOT_TOKEN, TELEGRAM_CHAT_ID
Email (Resend)	RESEND_API_KEY
Discord	DISCORD_WEBHOOK_URL
Slack	SLACK_WEBHOOK_URL
Generic webhook	your URL

Setup for each (BotFather, Resend domain verification, webhooks) + the generic payload shape: docs/notifications.md.

🔐 Access control

Four roles — SUPER_ADMIN · ADMIN · CLIENT · VIEWER — configured under access: in blip.config.yaml and enforced by the Worker with password auth at the edge. Clients see only their own groups and sites. A public status page (public sites only) can be served without login.

🔒 Details, principal/password setup, and how to put Cloudflare Access in front of the admin pages: docs/access-control.md.

📊 How it works

1. Probe — every 5 min, the Worker's scheduled() handler checks each site.
2. Store — results are appended to Cloudflare D1 and precomputed into JSON blobs.
3. Notify — state changes fan out to your channels.
4. Serve — the same Worker's fetch() serves the dashboard SPA and the /data/*.json API from D1.
5. View — users hit a fast status page + admin dashboard, behind RBAC.

Shared types in packages/shared/src/types.ts keep the Worker (writer) and dashboard (reader) in lockstep.

📈 Scaling limits

The 5-minute cron and Cloudflare's free tier set practical limits. Each tick probes all sites within the Worker's CPU budget; slower SSL/domain probes are cached and refreshed only every few hours:

Sites	Recommended interval
≤ 50	5 min
≤ 100	10 min

Need more? Raise the interval, or split across multiple Workers. The free tier covers 100k Worker requests/day — far more than a 5-minute cron consumes.

🆚 Blip vs Upptime vs Uptime Kuma

	Blip	Upptime	Uptime Kuma
Hosting model	One Cloudflare Worker + D1	GitHub Actions + Pages	Self-hosted server
Needs a server	❌	❌	✅
Cost	$0	$0	Server cost
Monitor LAN/homelab (no open ports)	✅ (Tunnel + Access)	❌	⚠️ (needs the server on-LAN)
Multi-tenant groups	✅	❌	⚠️
RBAC (roles)	✅ (4 roles)	❌	⚠️ single admin
Client-scoped dashboards	✅	❌	❌
HTTP / TCP / SSL / domain checks	✅	⚠️	✅
JSON-body assertions	✅	⚠️	⚠️
Status page	✅	✅	✅

⚠️ = partial / via plugins or workarounds. Both Upptime and Uptime Kuma are great projects — pick Blip when you want serverless and multi-tenant RBAC and homelab monitoring without exposing ports.

🗺️ Roadmap

• HTTP / TCP / SSL / domain checks
• Keyword + JSON-body assertions
• Multi-tenant groups + RBAC
• Telegram / Email / Discord / Slack / webhook channels
• Interactive setup wizard + demo seeder
• Maintenance windows (suppress alerts during planned work)
• Status-page subscriptions (email/RSS) for end users
• Per-incident public post-mortems
• Response-time SLO/percentile widgets
• More channels (Microsoft Teams, PagerDuty native, Opsgenie)
• Multi-region probes across Worker locations

Have an idea? Open a feature request.

🤝 Contributing

PRs and issues are very welcome! Start with CONTRIBUTING.md and our Code of Conduct. Good first steps: improve docs, add a notification channel, or tackle a roadmap item.

npm install
npm run typecheck && npm test && npm run build   # what CI runs

📄 License

MIT © Blip contributors. White-label it, ship it to clients, make it yours.

If Blip saves you a monitoring bill, give it a ⭐ — it helps a lot.