docker-compose.yml

# =============================================================================
# JobHunter (Arachnode) — Production Docker Compose
# =============================================================================
#
# REQUIRED ENVIRONMENT VARIABLES — copy .env.example to .env before running:
#
#   JOBHUNTER_API_KEY        API key for dashboard authentication (set a strong random string)
#   POSTGRES_USER            Database user (default: jobuser)
#   POSTGRES_PASSWORD        Database password (default: jobpass)
#   POSTGRES_DB              Database name (default: jobsdb)
#   GITHUB_TOKEN             GitHub personal access token for contact discovery
#   GMAIL_ADDRESS            Gmail address for sending cold emails
#   GMAIL_APP_PASSWORD       Gmail App Password (NOT your account password)
#   YOUR_NAME                Your display name for cold email templates
#   YOUR_GITHUB_URL          Your GitHub profile URL
#   JOBSEEKER_ROLE           Target job title (default: Backend Engineer)
#   JOBSEEKER_STACK          Comma-separated tech stack (default: Python,FastAPI,PostgreSQL,Redis,Go)
#   OLLAMA_BASE_URL          Ollama endpoint for LLM (default: http://host.docker.internal:11434)
#   GATEWAY_PORT             Host port for the dashboard (default: 8080)
#
# QUICK START:
#   cp .env.example .env    # then edit with your values
#   docker compose up -d
#   open http://localhost:8080
#
# Postgres and Redis are intentionally NOT exposed to the host.
# Use docker exec for direct access during development:
#   docker exec -it $(docker compose ps -q postgres) psql -U jobuser jobsdb
#   docker exec -it $(docker compose ps -q redis) redis-cli

# ---------------------------------------------------------------------------
# DNS configuration for reliable pip installs during build
# ---------------------------------------------------------------------------
x-build-dns: &build-dns
  network: host

services:

  # ── Infrastructure ──────────────────────────────────────────────────────────

  postgres:
    image: postgres:16-alpine
    environment:
      POSTGRES_USER:     ${POSTGRES_USER:-jobuser}
      POSTGRES_PASSWORD: ${POSTGRES_PASSWORD:-jobpass}
      POSTGRES_DB:       ${POSTGRES_DB:-jobsdb}
    volumes:
      - pgdata:/var/lib/postgresql/data
    # No ports exposed — access via: docker exec -it <container> psql -U jobuser jobsdb
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U ${POSTGRES_USER:-jobuser} -d ${POSTGRES_DB:-jobsdb}"]
      interval: 5s
      timeout: 5s
      retries: 10
    networks: [jobnet]
    restart: unless-stopped

  redis:
    image: redis:7-alpine
    command: redis-server --maxmemory 512mb --maxmemory-policy allkeys-lru --appendonly yes
    volumes:
      - redis_data:/data
    # No ports exposed — access via: docker exec -it <container> redis-cli
    healthcheck:
      test: ["CMD", "redis-cli", "ping"]
      interval: 5s
      timeout: 3s
      retries: 5
    networks: [jobnet]
    restart: unless-stopped

  # ── Application Services ────────────────────────────────────────────────────

  crawler:
    build:
      context: ./crawler-service
      dockerfile: Dockerfile
      <<: *build-dns
    environment:
      REDIS_HOST: redis
      REDIS_PORT: "6379"
      JOBSEEKER_ROLE:  ${JOBSEEKER_ROLE:-Backend Engineer}
      JOBSEEKER_STACK: ${JOBSEEKER_STACK:-Python,FastAPI,PostgreSQL,Redis,Go}
    depends_on:
      redis:
        condition: service_healthy
    restart: on-failure
    networks: [jobnet]
    healthcheck:
      test: ["CMD-SHELL", "curl -sf http://localhost:8004/health || exit 1"]
      interval: 15s
      timeout: 5s
      retries: 5
      start_period: 30s

  aggregator:
    build:
      context: ./aggregator-service
      dockerfile: Dockerfile
      <<: *build-dns
    environment:
      REDIS_HOST: redis
      REDIS_PORT: "6379"
      DATABASE_URL: postgresql://${POSTGRES_USER:-jobuser}:${POSTGRES_PASSWORD:-jobpass}@postgres:5432/${POSTGRES_DB:-jobsdb}
    depends_on:
      redis:
        condition: service_healthy
      postgres:
        condition: service_healthy
    networks: [jobnet]
    restart: on-failure
    healthcheck:
      test: ["CMD-SHELL", "curl -sf http://localhost:8000/health || exit 1"]
      interval: 10s
      timeout: 5s
      retries: 5
      start_period: 20s

  scraper:
    build:
      context: ./scraper-service
      dockerfile: Dockerfile
      <<: *build-dns
    environment:
      REDIS_HOST: redis
      REDIS_PORT: "6379"
      DATABASE_URL: postgresql://${POSTGRES_USER:-jobuser}:${POSTGRES_PASSWORD:-jobpass}@postgres:5432/${POSTGRES_DB:-jobsdb}
    depends_on:
      redis:
        condition: service_healthy
    networks: [jobnet]
    restart: on-failure
    deploy:
      resources:
        limits:
          memory: 1G
    healthcheck:
      test: ["CMD-SHELL", "curl -sf http://localhost:8000/health || exit 1"]
      interval: 10s
      timeout: 5s
      retries: 5
      start_period: 30s

  contact:
    build:
      context: ./contact-discovery-service
      dockerfile: Dockerfile
      <<: *build-dns
    environment:
      DATABASE_URL: postgresql://${POSTGRES_USER:-jobuser}:${POSTGRES_PASSWORD:-jobpass}@postgres:5432/${POSTGRES_DB:-jobsdb}
      GITHUB_TOKEN: ${GITHUB_TOKEN:-}
      REDIS_HOST: redis
      REDIS_PORT: "6379"
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
    networks: [jobnet]
    restart: on-failure
    deploy:
      resources:
        limits:
          memory: 1G
    healthcheck:
      test: ["CMD-SHELL", "curl -sf http://localhost:8000/health || exit 1"]
      interval: 10s
      timeout: 5s
      retries: 5
      start_period: 45s

  email-gen:
    build:
      context: ./email-generator-service
      dockerfile: Dockerfile
      <<: *build-dns
    environment:
      DATABASE_URL: postgresql://${POSTGRES_USER:-jobuser}:${POSTGRES_PASSWORD:-jobpass}@postgres:5432/${POSTGRES_DB:-jobsdb}
      OLLAMA_BASE_URL:    ${OLLAMA_BASE_URL:-http://host.docker.internal:11434}
      GMAIL_ADDRESS:      ${GMAIL_ADDRESS:-}
      GMAIL_APP_PASSWORD: ${GMAIL_APP_PASSWORD:-}
      YOUR_NAME:          ${YOUR_NAME:-Applicant}
      YOUR_GITHUB_URL:    ${YOUR_GITHUB_URL:-}
    depends_on:
      postgres:
        condition: service_healthy
    networks: [jobnet]
    restart: on-failure
    healthcheck:
      test: ["CMD-SHELL", "curl -sf http://localhost:8000/health || exit 1"]
      interval: 10s
      timeout: 5s
      retries: 5
      start_period: 20s

  agent-search:
    build:
      context: ./agent-search-service
      dockerfile: Dockerfile
      <<: *build-dns
    environment:
      DATABASE_URL: postgresql://${POSTGRES_USER:-jobuser}:${POSTGRES_PASSWORD:-jobpass}@postgres:5432/${POSTGRES_DB:-jobsdb}
      REDIS_HOST: redis
      REDIS_PORT: "6379"
      AGENT_PROVIDER: ${AGENT_PROVIDER:-ollama}
      AGENT_MODEL: ${AGENT_MODEL:-mistral}
      OLLAMA_BASE_URL: ${OLLAMA_BASE_URL:-http://host.docker.internal:11434}
      OPENAI_API_KEY: ${OPENAI_API_KEY:-}
      ANTHROPIC_API_KEY: ${ANTHROPIC_API_KEY:-}
    depends_on:
      postgres:
        condition: service_healthy
      redis:
        condition: service_healthy
    networks: [jobnet]
    restart: on-failure
    deploy:
      resources:
        limits:
          memory: 512M
    healthcheck:
      test: ["CMD-SHELL", "curl -sf http://localhost:8009/health || exit 1"]
      interval: 15s
      timeout: 5s
      retries: 5
      start_period: 20s

  gemini-agent:
    build:
      context: ./gemini-agent-service
      dockerfile: Dockerfile
      <<: *build-dns
    environment:
      DATABASE_URL: postgresql://${POSTGRES_USER:-jobuser}:${POSTGRES_PASSWORD:-jobpass}@postgres:5432/${POSTGRES_DB:-jobsdb}
      GEMINI_API_KEY: ${GEMINI_API_KEY:-}
      AGGREGATOR_URL: http://aggregator:8000
      SCRAPER_URL: http://scraper:8000
    depends_on:
      postgres:
        condition: service_healthy
    networks: [jobnet]
    restart: on-failure
    healthcheck:
      test: ["CMD-SHELL", "curl -sf http://localhost:8010/health || exit 1"]
      interval: 15s
      timeout: 5s
      retries: 5
      start_period: 20s

  gateway:
    build:
      context: ./gateway
      dockerfile: Dockerfile
      <<: *build-dns
    ports:
      - "${GATEWAY_PORT:-8080}:8080"
    environment:
      AGGREGATOR_URL:    http://aggregator:8000
      SCRAPER_URL:       http://scraper:8000
      CONTACT_URL:       http://contact:8000
      EMAIL_GEN_URL:     http://email-gen:8000
      AGENT_SEARCH_URL:  http://agent-search:8009
      GEMINI_AGENT_URL:  http://gemini-agent:8010
      CRAWLER_URL:       http://crawler:8004
      SUMMARY_PATH:      /data/run_summary.json
      JOBHUNTER_API_KEY: ${JOBHUNTER_API_KEY:-}
    volumes:
      - run_summary:/data:ro
    depends_on:
      aggregator:
        condition: service_healthy
      scraper:
        condition: service_healthy
      contact:
        condition: service_healthy
      email-gen:
        condition: service_healthy
      agent-search:
        condition: service_healthy
    networks: [jobnet]
    restart: on-failure
    healthcheck:
      test: ["CMD-SHELL", "curl -sf http://localhost:8080/api/health || exit 1"]
      interval: 15s
      timeout: 10s
      retries: 5
      start_period: 30s

  scheduler:
    build:
      context: ./scheduler
      dockerfile: Dockerfile
      <<: *build-dns
    environment:
      GATEWAY_URL:             http://gateway:8080
      CRAWLER_URL:             http://crawler:8004
      JOBHUNTER_API_KEY:       ${JOBHUNTER_API_KEY:-}
      JOBSEEKER_ROLE:          ${JOBSEEKER_ROLE:-Backend Engineer}
      JOBSEEKER_STACK:         ${JOBSEEKER_STACK:-Python,FastAPI,PostgreSQL,Redis,Go}
      CRAWL_INTERVAL_HOURS:    ${CRAWL_INTERVAL_HOURS:-8}
      DISCOVER_INTERVAL_HOURS: ${DISCOVER_INTERVAL_HOURS:-24}
      SCRAPER_WAIT_SECS:       ${SCRAPER_WAIT_SECS:-60}
      DISCOVER_DELAY_SECS:     ${DISCOVER_DELAY_SECS:-30}
      SUMMARY_PATH:            /data/run_summary.json
    volumes:
      - run_summary:/data
    depends_on:
      gateway:
        condition: service_healthy
      crawler:
        condition: service_healthy
    networks: [jobnet]
    restart: on-failure

# ---------------------------------------------------------------------------
# Named volumes
# ---------------------------------------------------------------------------
volumes:
  pgdata:
    driver: local
  redis_data:
    driver: local
  run_summary:
    driver: local

# ---------------------------------------------------------------------------
# Networks
# ---------------------------------------------------------------------------
networks:
  jobnet:
    driver: bridge
    driver_opts:
      com.docker.network.bridge.name: jobnet0