Generate Checker Config

You are helping the user generate a config.yaml for Checker — an open-source distributed health-checking and alerting system. Use everything you know about the user's infrastructure from this conversation to produce a complete, ready-to-use configuration.

How to use this prompt

Paste this file into a Claude conversation where you've already discussed your infrastructure, or describe your setup and ask Claude to generate the config.

What you should ask the user (if not already known)

1. What services/endpoints do you want to monitor? (URLs, hosts, ports, databases)
2. Where should alerts go? (Telegram, Slack, Discord, PagerDuty, OpsGenie, Teams, Email, ntfy)
3. Database preference? PostgreSQL (production) or SQLite (dev/demo)
4. Authentication? OIDC, password, API keys, or none
5. Multi-region? If yes, which regions and consensus settings

Then generate a complete config.yaml following the reference below.

---

Config Structure

# ──────────────────────────────────────────────
# Server
# ──────────────────────────────────────────────
server_port: "8080"

# ──────────────────────────────────────────────
# Database
# ──────────────────────────────────────────────
db:
  driver: postgres          # "postgres" or "sqlite"
  # PostgreSQL:
  host: localhost
  username: checker
  password: ""
  database: checker
  # database_url: postgres://user:pass@host:5432/checker  # alternative
  # SQLite:
  # driver: sqlite
  # dsn: ./checker.db

# ──────────────────────────────────────────────
# Defaults (applied to all checks unless overridden)
# ──────────────────────────────────────────────
defaults:
  duration: 30s             # Check interval
  alerts_channel: telegram  # Default alert channel name (must match a key in alerts:)

# ──────────────────────────────────────────────
# Alert Channels
# ──────────────────────────────────────────────
alerts:
  # Each key is a channel name you reference from checks.
  # Include ONLY the channels the user actually needs.

  telegram:
    type: telegram
    bot_token: "BOT_TOKEN"
    critical_channel: "CHAT_ID"
    noncritical_channel: "CHAT_ID"

  slack:
    type: slack
    webhook_url: "https://hooks.slack.com/services/T.../B.../xxx"

  discord:
    type: discord
    bot_token: "BOT_TOKEN"
    app_id: "APP_ID"
    default_channel: "CHANNEL_ID"

  pagerduty:
    type: pagerduty
    routing_key: "ROUTING_KEY"

  opsgenie:
    type: opsgenie
    api_key: "API_KEY"
    region: "us"              # "us" or "eu"

  email:
    type: email
    smtp_host: "smtp.example.com"
    smtp_port: 587
    smtp_user: "alerts@example.com"
    smtp_password: "PASSWORD"
    from: "alerts@example.com"
    to: ["oncall@example.com"]
    use_tls: true

  teams:
    type: teams
    webhook_url: "https://outlook.webhook.office.com/webhookb2/..."

  ntfy:
    type: ntfy
    topic: "checker-alerts"
    server_url: "https://ntfy.sh"   # default; use your own server if self-hosted
    # token: ""                     # optional auth
    # click_url: "https://checker.example.com"

# ──────────────────────────────────────────────
# Bot Integrations (optional — for interactive commands)
# ──────────────────────────────────────────────
# slack_app:
#   bot_token: "xoxb-..."
#   signing_secret: "SECRET"
#   default_channel: "C123456"

# telegram_app:
#   bot_token: "123456:ABCdef..."
#   secret_token: "SECRET"
#   default_chat_id: "CHAT_ID"
#   webhook_url: "https://checker.example.com/telegram/webhook"

# discord_app:
#   bot_token: "BOT_TOKEN"
#   app_id: "APP_ID"
#   public_key: "PUBLIC_KEY"
#   default_channel: "CHANNEL_ID"

# ──────────────────────────────────────────────
# Authentication (optional)
# ──────────────────────────────────────────────
# auth:
#   password: "your-password"
#   # OR OIDC:
#   oidc:
#     issuer_url: "https://auth.example.com"
#     client_id: "CLIENT_ID"
#     client_secret: "CLIENT_SECRET"
#     redirect_url: "https://checker.example.com/auth/callback"
#   api_keys: ["key1", "key2"]

# ──────────────────────────────────────────────
# Multi-Region Consensus (optional)
# ──────────────────────────────────────────────
# consensus:
#   region: "us-east-1"
#   min_regions: 2
#   evaluation_interval: "10s"
#   timeout: "30s"

# ──────────────────────────────────────────────
# Projects & Checks
# ──────────────────────────────────────────────
projects:
  # Group checks into projects. Each project can override defaults.
  my_project:
    healthchecks:
      web:
        checks:
          api_health:
            type: http
            url: "https://api.example.com/health"
            # ... check-specific fields (see reference below)

---

Check Types Reference

Generate checks using the types below. Only include fields that are relevant.

HTTP

type: http
url: "https://example.com/health"
timeout: "10s"
code: [200]                           # Expected status codes (default: [200])
answer: "ok"                          # Expected substring in response body
answer_present: true                  # true = body must contain answer; false = must NOT contain
headers: ["Authorization: Bearer TOKEN"]
cookies: ["session=abc"]
skip_check_ssl: false                 # Skip TLS verification
ssl_expiration_period: "30d"          # Alert if SSL cert expires within this period
stop_follow_redirects: false
auth:
  user: "username"
  password: "password"

TCP

type: tcp
host: "db.example.com"
port: 5432
timeout: "10s"

ICMP (Ping)

type: icmp
host: "server.example.com"
count: 4                              # Ping count
timeout: "10s"

DNS

type: dns
domain: "example.com"
record_type: "A"                      # A, AAAA, MX, TXT, NS, CNAME
host: "8.8.8.8"                      # Custom resolver (optional)
expected: "93.184.216.34"            # Expected value (optional)
timeout: "10s"

SSH

type: ssh
host: "server.example.com"
port: 22
timeout: "10s"
expect_banner: "SSH-2.0"             # Expected banner prefix (optional)

SSL Certificate

type: ssl_cert
host: "example.com"
port: 443
timeout: "10s"
expiry_warning_days: 30
validate_chain: true

SMTP

type: smtp
host: "mail.example.com"
port: 587
timeout: "10s"
starttls: true

Domain Expiry (WHOIS)

type: domain_expiry
domain: "example.com"
expiry_warning_days: 30
timeout: "10s"

PostgreSQL

# Simple query check
type: pgsql_query
host: "db.example.com"
port: 5432
timeout: "10s"
pgsql:
  username: "monitor"
  password: "PASSWORD"
  dbname: "mydb"
  sslmode: "prefer"                   # disable, allow, prefer, require
  query: "SELECT 1"
  response: "1"

# Replication lag
type: pgsql_replication
pgsql:
  username: "monitor"
  password: "PASSWORD"
  dbname: "mydb"
  lag: "10s"
  server_list: ["replica1:5432", "replica2:5432"]

# Timestamp freshness
type: pgsql_query_unixtime
pgsql:
  query: "SELECT EXTRACT(EPOCH FROM updated_at)::int FROM jobs ORDER BY updated_at DESC LIMIT 1"
  difference: "300s"                  # Alert if older than 5 minutes

MySQL

# Simple query check
type: mysql_query
host: "db.example.com"
port: 3306
timeout: "10s"
mysql:
  username: "monitor"
  password: "PASSWORD"
  dbname: "mydb"
  query: "SELECT 1"
  response: "1"

# Replication lag
type: mysql_replication
mysql:
  username: "monitor"
  password: "PASSWORD"
  lag: "10s"
  server_list: ["replica1:3306"]

Redis

type: redis
host: "redis.example.com"
port: 6379
timeout: "10s"
password: "PASSWORD"                  # optional
db: 0

MongoDB

type: mongodb
mongodb_uri: "mongodb://user:pass@host:27017/dbname"
timeout: "10s"

gRPC Health

type: grpc_health
host: "service.example.com:50051"
timeout: "10s"
use_tls: false

WebSocket

type: websocket
url: "ws://example.com/ws"
timeout: "10s"

Passive (Heartbeat)

type: passive
timeout: "60s"                        # Alert if no ping received within this period
# Heartbeat endpoint: POST /api/checks/{uuid}/ping

---

Common Check Options

Every check supports these optional fields:

duration: "30s"                       # Check interval (overrides project/global default)
timeout: "10s"                        # Check timeout
severity: "critical"                  # "critical", "warning", "info"
enabled: true
retry_count: 2                        # Retry N times before marking as failed
retry_interval: "5s"                  # Wait between retries
re_alert_interval: "30m"             # Re-send alert every N if still failing
alert_channels: ["telegram", "slack"] # Send alerts to these channels
escalation_policy_name: "on_call"    # Escalation policy name

---

Generation Guidelines

When generating the config:

1. Group checks by project — use meaningful project names (e.g., "Production API", "Databases", "Infrastructure")
2. Set appropriate intervals — critical services: 15-30s, standard: 1-5m, SSL/domain: 1h-24h
3. Layer alerting — critical checks to PagerDuty/OpsGenie + Slack; warnings to Slack/Telegram only
4. Add SSL checks for every HTTPS endpoint with ssl_expiration_period: "30d"
5. Add DNS checks for critical domains
6. Add domain expiry checks for owned domains
7. Use retries for flaky checks: retry_count: 2, retry_interval: "5s"
8. Set re_alert_interval to avoid alert fatigue: "30m" for critical, "2h" for warnings
9. Replace all placeholder values (BOT_TOKEN, PASSWORD, etc.) with comments showing what's needed
10. Only include alert channels the user mentioned — don't add unused ones

Environment Variables

Sensitive values can be set via environment variables instead of hardcoding in YAML:

DATABASE_URL=postgres://user:pass@host:5432/checker
SLACK_BOT_TOKEN=xoxb-...
TELEGRAM_APP_BOT_TOKEN=123456:ABCdef...
AUTH_PASSWORD=secure-password

Running Checker

# With config file
./checker -config config.yaml

# With Docker
docker run -v $(pwd)/config.yaml:/config.yaml ghcr.io/ensafely/checker -config /config.yaml

# Debug mode
./checker -config config.yaml --debug