docker-compose.yml

name: ccam

# The dashboard is intentionally a single writer. CCAM stores its operational
# state in one SQLite database, so never scale agent-monitor above one replica
# or run two active containers against the same data volume.
services:
  agent-monitor:
    image: ${CCAM_IMAGE:-ccam-dashboard:2.4.5}
    build:
      context: .
      dockerfile: Dockerfile
      target: ${CCAM_DOCKER_TARGET:-runtime}
    restart: unless-stopped
    stop_grace_period: 30s
    ports:
      - "${CCAM_DASHBOARD_BIND:-127.0.0.1}:${DASHBOARD_PORT:-4820}:4820"
    environment:
      NODE_ENV: production
      DASHBOARD_HOST: 0.0.0.0
      DASHBOARD_PORT: 4820
      DASHBOARD_DATA_DIR: /app/data
      DASHBOARD_ENV_PATH: /app/config/.env
      DASHBOARD_LIVENESS_PROBE: 0
      CLAUDE_HOME: /home/node/.claude
      DASHBOARD_CURSOR_HOME: /home/node/.cursor
      DASHBOARD_CODEX_HOME: /home/node/.codex
      DASHBOARD_ALLOWED_HOSTS: ${DASHBOARD_ALLOWED_HOSTS:-localhost,agent-monitor,ccam-nginx}
      DASHBOARD_TOKEN: ${DASHBOARD_TOKEN:-}
      DASHBOARD_TOKEN_FILE: /run/ccam-secrets/dashboard-token
      DASHBOARD_HOOK_TOKEN: ${DASHBOARD_HOOK_TOKEN:-}
      DASHBOARD_HOOK_TOKEN_FILE: /run/ccam-secrets/hook-token
      TZ: ${TZ:-UTC}
    volumes:
      - ${CCAM_DATA_VOLUME:-ccam-data}:/app/data
      - ${CCAM_CONFIG_VOLUME:-ccam-config}:/app/config
      - ${CLAUDE_HOME:-~/.claude}:/home/node/.claude:${CCAM_AGENT_HOME_MODE:-ro}
      - ${CURSOR_HOME:-~/.cursor}:/home/node/.cursor:${CCAM_AGENT_HOME_MODE:-ro}
      - ${CODEX_HOME:-~/.codex}:/home/node/.codex:${CCAM_AGENT_HOME_MODE:-ro}
      - ${CCAM_WORKSPACE_DIR:-.}:/workspace:${CCAM_WORKSPACE_MODE:-ro}
      - ${CCAM_SECRETS_DIR:-./deployments/secrets}:/run/ccam-secrets:ro
    tmpfs:
      - /tmp:size=512m,mode=1777
    read_only: true
    cap_drop:
      - ALL
    security_opt:
      - no-new-privileges:true
    healthcheck:
      test:
        [
          "CMD",
          "node",
          "-e",
          "fetch('http://127.0.0.1:4820/api/health').then((response)=>process.exit(response.ok?0:1)).catch(()=>process.exit(1))",
        ]
      interval: 15s
      timeout: 5s
      retries: 5
      start_period: 20s
    networks:
      - ccam

  # Optional internal MCP service. It reaches the dashboard through the private
  # Compose network and publishes MCP on host loopback only.
  mcp:
    profiles: ["mcp"]
    image: ${CCAM_MCP_IMAGE:-ccam-mcp:2.4.5}
    build:
      context: .
      dockerfile: mcp/Dockerfile
      target: runtime
    restart: unless-stopped
    ports:
      - "${CCAM_MCP_BIND:-127.0.0.1}:${MCP_HTTP_PORT:-8819}:8819"
    depends_on:
      agent-monitor:
        condition: service_healthy
    environment:
      MCP_TRANSPORT: http
      MCP_HTTP_HOST: 0.0.0.0
      MCP_HTTP_PORT: 8819
      MCP_HTTP_AUTH_TOKEN_FILE: /run/ccam-secrets/mcp-token
      MCP_DASHBOARD_BASE_URL: http://agent-monitor:4820
      MCP_DASHBOARD_API_TOKEN: ${DASHBOARD_TOKEN:-}
      MCP_DASHBOARD_API_TOKEN_FILE: /run/ccam-secrets/dashboard-token
      MCP_DASHBOARD_ALLOW_MUTATIONS: ${MCP_DASHBOARD_ALLOW_MUTATIONS:-false}
      MCP_DASHBOARD_ALLOW_DESTRUCTIVE: ${MCP_DASHBOARD_ALLOW_DESTRUCTIVE:-false}
      MCP_LOG_LEVEL: ${MCP_LOG_LEVEL:-info}
    volumes:
      - ${CCAM_SECRETS_DIR:-./deployments/secrets}:/run/ccam-secrets:ro
    cap_drop:
      - ALL
    read_only: true
    tmpfs:
      - /tmp:size=64m,mode=1777
    security_opt:
      - no-new-privileges:true
    networks:
      - ccam

  # Optional cloud/LAN edge. TLS should terminate at the cloud load balancer or
  # at a mounted provider-specific Nginx TLS configuration.
  nginx:
    profiles: ["edge"]
    image: nginxinc/nginx-unprivileged:1.30.0-alpine@sha256:808f7846d21a9c94cf53833e8807a00a33fd0b65cc47fb05b79efe366c2d201f
    entrypoint: ["/bin/sh", "/etc/nginx/ccam-entrypoint.sh"]
    restart: unless-stopped
    depends_on:
      agent-monitor:
        condition: service_healthy
    ports:
      - "${CCAM_EDGE_BIND:-127.0.0.1}:${CCAM_EDGE_PORT:-8080}:8080"
    volumes:
      - ./deployments/nginx/nginx.conf:/etc/nginx/nginx.conf.template:ro
      - ./deployments/nginx/entrypoint.sh:/etc/nginx/ccam-entrypoint.sh:ro
      - ./deployments/nginx/conf.d:/etc/nginx/conf.d:ro
      - ${CCAM_NGINX_HOOK_POLICY:-./deployments/nginx/snippets/hooks-blocked.conf}:/etc/nginx/snippets/hooks-policy.conf:ro
      - ${CCAM_NGINX_MCP_POLICY:-./deployments/nginx/snippets/mcp-blocked.conf}:/etc/nginx/snippets/mcp-policy.conf:ro
    read_only: true
    tmpfs:
      - /tmp:size=64m,mode=1777
      - /var/cache/nginx:size=64m,mode=0755
    cap_drop:
      - ALL
    security_opt:
      - no-new-privileges:true
    healthcheck:
      test: ["CMD", "wget", "-qO-", "http://127.0.0.1:8080/healthz"]
      interval: 15s
      timeout: 5s
      retries: 5
      start_period: 10s
    networks:
      - ccam

networks:
  ccam:
    name: ccam

volumes:
  ccam-data:
    name: ccam-data
  ccam-config:
    name: ccam-config