agentleFS
joshuadavidthomasjoshuadavidthomas/agent-skills/coolify-compose/SKILL.md

coolify-compose

Skill49 starsChanged 6 months agoMITView on GitHub

Convert Docker Compose files to Coolify templates. Use when creating Coolify services, converting docker-compose.yml for Coolify deployment, working with SERVICE_URL/SERVICE_PASSWORD magic variables, or troubleshooting Coolify compose errors.

  • Reads credentials

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 4 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Every Coolify template needs a header and magic variables:

    line 73
  2. Very specific:

    When converting a standard docker-compose.yml:

    line 112
  3. Very specific:

    If the compose has build: entries:

    line 116
  4. Distinctive:

    Repository mode: Keep them. Coolify will build from the Dockerfile.

    line 117
  5. Distinctive:

    ⚠️ Important: Use hyphens, not underscores, before port numbers:

    line 243
  6. Distinctive:

    When converting a compose file without an official template, analyze it and use the matching example:

    line 370
  7. Very specific:

    Look for postgres, mysql, mariadb, mongo images — if 1 database, use with-database/

    line 380
  8. Very specific:

    Look for mounted .xml, .json, .yml config files — if present, use multi-service/

    line 381
  9. Very specific:

    Look for clickhouse, redis, multiple databases — use multi-service/

    line 382

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Coolify Docker Compose

Convert standard Docker Compose files into Coolify-compatible templates with automatic credential generation, dynamic URLs, and one-click deployment.

Two Deployment Modes

Coolify supports two ways to deploy compose files with different capabilities:

1. Raw Compose (Paste Content)

Paste compose YAML directly into Coolify's UI. Limited feature set:

FeatureSupported
image:✅ Yes
build:❌ No - must use pre-built images
External config files❌ No - must use inline content:
YAML anchors (&, *)✅ Yes - resolved by YAML parser
Coolify magic variables✅ Yes
content: for inline files✅ Yes

Use when: Quick deployments, simple services, no custom images needed.

2. Repository Mode (Git URL)

Point Coolify to a Git repository containing your compose file. Full Docker Compose features:

FeatureSupported
image:✅ Yes
build:✅ Yes - builds from Dockerfile in repo
External config files✅ Yes - relative paths work
Coolify magic variables✅ Yes
content: for inline files✅ Yes

Use when: Custom images needed, complex multi-file setups, existing docker-compose.yml in a repo.

Repository setup:

my-service/
├── compose.yml          # or docker-compose.yml
├── custom-image/
│   ├── Dockerfile
│   └── config.sql
└── other-files/
# compose.yml - can use build:
services:
  app:
    build:
      context: ./custom-image
      dockerfile: Dockerfile

Which Mode to Use?

Original compose has...Recommended mode
Only image: referencesEither works
build: directivesRepository mode
External config files to mountRepository mode (or use content: in raw)
Single simple serviceRaw mode is faster

Quick Start

Every Coolify template needs a header and magic variables:

# documentation: https://example.com/docs
# slogan: Brief description of the service
# category: backend
# tags: api, database, docker
# logo: svgs/myservice.svg
# port: 3000

services:
  app:
    image: myapp:latest
    environment:
      - SERVICE_URL_APP_3000           # Generates URL, routes proxy to port 3000
      - DATABASE_URL=postgres://${SERVICE_USER_POSTGRES}:${SERVICE_PASSWORD_POSTGRES}@db:5432/mydb
    depends_on:
      db:
        condition: service_healthy
    healthcheck:
      test: ["CMD", "wget", "-q", "--spider", "http://localhost:3000/health"]
      interval: 5s
      timeout: 10s
      retries: 10

  db:
    image: postgres:16-alpine
    environment:
      - POSTGRES_USER=$SERVICE_USER_POSTGRES
      - POSTGRES_PASSWORD=$SERVICE_PASSWORD_POSTGRES
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U $${POSTGRES_USER}"]
      interval: 5s
      timeout: 10s
      retries: 10

Conversion Checklist

When converting a standard docker-compose.yml:

0. Check for build: Directives

If the compose has build: entries:

  • Repository mode: Keep them. Coolify will build from the Dockerfile.
  • Raw mode: Replace with pre-built image: references or find equivalent images.
# Original with build:
services:
  custom-db:
    build: ./custom-postgres

# Raw mode: find or create pre-built image
  custom-db:
    image: your-registry.com/custom-postgres:latest

# Repository mode: keep build:, include Dockerfile in repo
  custom-db:
    build:
      context: ./custom-postgres
      dockerfile: Dockerfile

1. Add Header Metadata

# documentation: https://...    # Required: URL to official docs
# slogan: ...                   # Required: One-line description  
# category: ...                 # Required: backend, cms, monitoring, etc.
# tags: ...                     # Required: Comma-separated search terms
# logo: svgs/....svg            # Required: Path in Coolify's svgs/ folder
# port: ...                     # Recommended: Main service port

2. Replace Hardcoded Credentials

# ❌ Before
POSTGRES_PASSWORD=mysecretpassword
POSTGRES_USER=admin

# ✅ After  
POSTGRES_PASSWORD=$SERVICE_PASSWORD_POSTGRES
POSTGRES_USER=$SERVICE_USER_POSTGRES

3. Replace URLs with Magic Variables

# ❌ Before
APP_URL=https://myapp.example.com

# ✅ After
- SERVICE_URL_APP_3000    # Declares URL + proxy routing
- APP_URL=$SERVICE_URL_APP  # References it

4. Remove ports: for Proxied Services

Coolify's Traefik proxy handles routing. Only keep ports: for SSH, UDP, or proxy bypass.

# ❌ Before
ports:
  - "3000:3000"

# ✅ After
environment:
  - SERVICE_URL_APP_3000  # Proxy routes to container port 3000
# No ports: needed

5. Add Health Checks

healthcheck:
  test: ["CMD", "curl", "-f", "http://localhost:3000/health"]
  interval: 5s
  timeout: 10s
  retries: 10

6. Use depends_on with Conditions

depends_on:
  db:
    condition: service_healthy

Magic Variables Reference

Coolify generates values using SERVICE_<TYPE>_<IDENTIFIER>:

TypeExampleResult
PASSWORDSERVICE_PASSWORD_DBRandom password
PASSWORD_64SERVICE_PASSWORD_64_KEY64-char password
USERSERVICE_USER_ADMINRandom 16-char string
BASE64_64SERVICE_BASE64_64_SECRET64-char random string
REALBASE64_64SERVICE_REALBASE64_64_JWTActual base64-encoded string
HEX_32SERVICE_HEX_32_KEY64-char hex string
URLSERVICE_URL_APP_3000https://app-uuid.example.com + proxy
FQDNSERVICE_FQDN_APPapp-uuid.example.com (no scheme, no port suffix)

Declaration vs Reference (Critical)

For SERVICE_URL, the port suffix configures proxy routing but is not part of the variable name:

# Declare WITH port suffix (configures proxy to route to port 3000)
- SERVICE_URL_MYAPP_3000

# Reference WITHOUT port suffix (gets the URL value)
- APP_URL=$SERVICE_URL_MYAPP
- WEBHOOK_URL=${SERVICE_URL_MYAPP}/webhooks

SERVICE_FQDN is automatically available when SERVICE_URL is declared — no separate declaration needed:

# This single declaration...
- SERVICE_URL_MYAPP_3000

# ...makes BOTH of these available:
- FULL_URL=$SERVICE_URL_MYAPP           # https://myapp-uuid.example.com
- HOSTNAME=${SERVICE_FQDN_MYAPP}        # myapp-uuid.example.com

⚠️ Important: Use hyphens, not underscores, before port numbers:

SERVICE_URL_MY_SERVICE_3000  # ❌ Breaks parsing
SERVICE_URL_MY-SERVICE_3000  # ✅ Works

See references/magic-variables.md for complete list.

Coolify-Specific Extensions

Create Directory

volumes:
  - type: bind
    source: ./data
    target: /app/data
    is_directory: true  # Coolify creates this

Create File with Content

Useful in raw mode when you can't reference external files. In repository mode, you can just mount files normally.

# Raw mode: embed file content inline
volumes:
  - type: bind
    source: ./config.json
    target: /app/config.json
    content: |
      {"key": "${SERVICE_PASSWORD_APP}"}

# Repository mode: reference actual file in repo
volumes:
  - ./config/settings.json:/app/config.json:ro

Exclude from Health Checks

For migration/init containers that exit after running:

services:
  migrate:
    command: ["npm", "run", "migrate"]
    exclude_from_hc: true

Common Patterns

Database Connection

environment:
  - DATABASE_URL=postgres://${SERVICE_USER_POSTGRES}:${SERVICE_PASSWORD_POSTGRES}@db:5432/${POSTGRES_DB:-myapp}

Shared Credentials

Same SERVICE_PASSWORD_* identifier = same value across all services:

services:
  app:
    environment:
      - DB_PASS=$SERVICE_PASSWORD_POSTGRES
  db:
    environment:
      - POSTGRES_PASSWORD=$SERVICE_PASSWORD_POSTGRES  # Same value

Multi-Service URLs

services:
  frontend:
    environment:
      - SERVICE_URL_FRONTEND_3000
      - API_URL=$SERVICE_URL_API
  api:
    environment:
      - SERVICE_URL_API_8080=/api  # Path suffix

Health Check Patterns

# HTTP
test: ["CMD", "wget", "--spider", "-q", "http://localhost:8080"]

# PostgreSQL
test: ["CMD-SHELL", "pg_isready -U $${POSTGRES_USER} -d $${POSTGRES_DB}"]

# MySQL/MariaDB  
test: ["CMD", "healthcheck.sh", "--connect", "--innodb_initialized"]

# Redis
test: ["CMD", "redis-cli", "ping"]

# Always pass (use sparingly)
test: ["CMD", "echo", "ok"]

Environment Variable Syntax

environment:
  - NODE_ENV=production           # Hardcoded, hidden from UI
  - API_KEY=${API_KEY}            # Editable in UI (empty)
  - LOG_LEVEL=${LOG_LEVEL:-info}  # Editable with default
  - SECRET=${SECRET:?}            # Required - blocks deploy if empty

Troubleshooting

ProblemSolution
"No Available Server" errorCheck docker ps for unhealthy containers; verify healthcheck passes
Variables not in Coolify UIUse ${VAR} syntax; hardcoded VAR=value won't appear
Magic variables not generatingCheck spelling; ensure SERVICE_ prefix; verify Coolify v4.0.0-beta.411+
Port routing brokenUse SERVICE_URL_NAME_PORT; avoid underscores before port; remove ports:

Examples

First, check for an official template: Many popular services have official Coolify templates at github.com/coollabsio/coolify/tree/main/templates/compose. If one exists, use it as the reference for correct patterns.

When converting a compose file without an official template, analyze it and use the matching example:

Compose file has...Use example
1 service, no databaseexamples/simple/
2 services: app + database (postgres/mysql/mariadb)examples/with-database/
3+ services, or mounted config files, or multiple databasesexamples/multi-service/

Quick analysis:

  • Count the services: — if just 1, use simple/
  • Look for postgres, mysql, mariadb, mongo images — if 1 database, use with-database/
  • Look for mounted .xml, .json, .yml config files — if present, use multi-service/
  • Look for clickhouse, redis, multiple databases — use multi-service/

References

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.