docker-compose.amd64.yml

services:
  # Neo4j (Legacy - replaced by NornicDB)
  # neo4j:
  #   image: neo4j:5.15-community
  #   container_name: neo4j_db
  #   ports:
  #     - "7475:7474"  # HTTP Browser UI
  #     - "7688:7687"  # Bolt protocol
  #   volumes:
  #     - ./data/neo4j:/data
  #     - ./logs/neo4j:/logs
  #     - ./data/neo4j/import:/var/lib/neo4j/import
  #   environment:
  #     - NEO4J_AUTH=neo4j/${NEO4J_PASSWORD:-password}
  #     - NEO4J_dbms_memory_pagecache_size=512M
  #     - NEO4J_dbms_memory_heap_initial__size=512M
  #     - NEO4J_dbms_memory_heap_max__size=2G
  #     - NEO4J_PLUGINS=["apoc"]
  #   restart: unless-stopped
  #   healthcheck:
  #     test: ["CMD-SHELL", "cypher-shell -u neo4j -p $${NEO4J_PASSWORD:-password} 'RETURN 1' || exit 1"]
  #     interval: 30s
  #     timeout: 10s
  #     retries: 3
  #     start_period: 40s
  #   networks:
  #     - mcp_network
  nornicdb:
    image: timothyswt/nornicdb-amd64-cuda-bge-heimdall:v1.0.0
    container_name: nornicdb
    ports:
      - "7474:7474"  # HTTP API & Browser UI
      - "7687:7687"  # Bolt protocol (Neo4j compatible)
    volumes:
      - ./data/movies-test:/data
      - ./data/movies-test/export:/data/export # For loading Mimir exports
    environment:
      - NORNICDB_DATA_DIR=/data
      - NORNICDB_HTTP_PORT=7474
      - NORNICDB_BOLT_PORT=7687
      # Embedding configuration - llama.cpp OpenAI-compatible API
      - NORNICDB_EMBEDDING_PROVIDER=local
      - NORNICDB_MODELS_DIR=/app/models
      # # - NORNICDB_EMBEDDING_API_URL=http://llama-server:8080
      # # - NORNICDB_EMBEDDING_API_KEY=not-needed
      # # - NORNICDB_EMBEDDING_MODEL=bge-m3
      # # Use /app/models for bundled models (bge-m3 + qwen heimdall)
      # # Volume mount ./data/nornicdb:/data shadows /data/models, so use /app/models
      # # - NORNICDB_MODELS_DIR=/app/models
      # # # Authentication (disabled by default for dev)
      # - NORNICDB_NO_AUTH=${NORNICDB_NO_AUTH:-true}
      # - NORNICDB_ADMIN_PASSWORD=${NEO4J_PASSWORD:-password}
      # - NORNICDB_TOPOLOGY_AUTO_INTEGRATION_ENABLED=true
      # # Load export data on startup (disabled - starting fresh)
      # - NORNICDB_LOAD_EXPORT=
      # # Badger memory tuning (reduce memory footprint)
      # - GOGC=50
      # - NORNICDB_GPU_ENABLED=${NORNICDB_GPU_ENABLED:-true}
      # - NORNICDB_GPU_DEVICE_ID=${NORNICDB_GPU_DEVICE_ID:-0}
      # - NORNICDB_EMBEDDING_CACHE_SIZE=0
    restart: unless-stopped
    healthcheck:
      test: ["CMD-SHELL", "wget --spider -q http://localhost:7474/health || exit 1"]
      interval: 30s
      timeout: 10s
      retries: 3
      start_period: 10s
    networks:
      - mcp_network
    # deploy:
    #   resources:
    #     reservations:
    #       devices:
    #         - driver: nvidia
    #           count: 1
    #           capabilities: [gpu, compute, utility]

  # copilot-api:
  #   image: timothyswt/copilot-api:latest
  #   container_name: copilot_api_server
  #   ports:
  #     - "4141:4141"  # Fixed: copilot-api listens on 4141, not 3000
  #   volumes:
  #     - ./copilot-data:/root/.local/share/copilot-api  # Persist GitHub token
  #   environment:
  #     - NODE_ENV=production
  #     # Remove PORT=3000, the app uses 4141 by default
  #   restart: unless-stopped
  #   healthcheck:
  #     # Use CMD-SHELL so shell operators (||) work and allow a proper HTTP probe
  #     test: ["CMD-SHELL", "wget --spider -q http://localhost:4141/ || exit 1"]
  #     interval: 30s
  #     timeout: 10s
  #     retries: 5
  #     start_period: 15s
  #   networks:
  #     - mcp_network

  # Ollama (Legacy - kept for reference, can be removed)
  # ollama:
  #   build:
  #     context: ./docker/ollama
  #     dockerfile: Dockerfile
  #     args:
  #       - EMBEDDING_MODEL=${MIMIR_EMBEDDINGS_MODEL:-bge-m3}
  #     tags:
  #       - mimir-ollama:${VERSION:-1.0.0}
  #       - mimir-ollama:latest
  #   image: mimir-ollama:${VERSION:-1.0.0}
  #   container_name: ollama_server
  #   ports:
  #     - "11434:11434"
  #   volumes:
  #     - type: bind
  #       source: ollama_models
  #       target: /root/.ollama
  #   environment:
  #     - OLLAMA_HOST=0.0.0.0:11434
  #     - OLLAMA_ORIGINS=*
  #   restart: unless-stopped
  #   healthcheck:
  #     test: ["CMD", "ollama", "list"]
  #     interval: 10s
  #     timeout: 5s
  #     retries: 5
  #     start_period: 30s
  #   networks:
  #     - mcp_network
  #   deploy:
  #     resources:
  #       reservations:
  #         devices:
  #           - driver: nvidia
  #             count: 1
  #             capabilities: [gpu]

  # llama.cpp server - OpenAI-compatible embeddings
  # llama-server:
  #   image: timothyswt/llama-cpp-server-amd64-mxbai-cuda:latest
  #   container_name: llama_server
  #   ports:
  #     - "11434:8080"  # External 11434 -> Internal 8080 (llama.cpp default)
  #   # Note: Model is bundled in the image (bge-m3, 1024 dims)
  #   # No volumes needed unless you want to add additional models
  #   restart: unless-stopped
  #   healthcheck:
  #     test: ["CMD-SHELL", "curl -f http://localhost:8080/health || exit 1"]
  #     interval: 30s
  #     timeout: 10s
  #     retries: 3
  #     start_period: 30s
  #   networks:
  #     - mcp_network
  #   deploy:
  #     resources:
  #       reservations:
  #         devices:
  #           - driver: nvidia
  #             count: 1
  #             capabilities: [gpu]

  # llama.cpp Vision Server - 2B model (faster, lower memory)
  # Qwen2.5-VL 2B with vision understanding and CUDA acceleration
  # Build with: npm run llama:build-qwen-2b-cuda
  # For use without cuda `image: timothyswt/llama-cpp-server-amd64-qwen-vl-2b:latest`
  # 
  # llama-vl-server-2b:
  #   image: timothyswt/llama-cpp-server-amd64-qwen-vl-2b-cuda:latest
  #   container_name: llama_server_vision_2b
  #   ports:
  #     - "8081:8080"  # Vision 2B API endpoint
  #   restart: unless-stopped
  #   healthcheck:
  #     test: ["CMD-SHELL", "curl -f http://localhost:8080/health || exit 1"]
  #     interval: 30s
  #     timeout: 10s
  #     retries: 3
  #     start_period: 60s  # Vision models take longer to load
  #   networks:
  #     - mcp_network
  #   deploy:
  #     resources:
  #       limits:
  #         memory: 8G  # 2B model + vision projector
  #       reservations:
  #         memory: 4G
  #         devices:
  #           - driver: nvidia
  #             count: 1
  #             capabilities: [gpu]

  # llama.cpp Vision Server - 7B model (higher quality, more memory)
  # Qwen2.5-VL 7B with vision understanding and CUDA acceleration
  # Build with: npm run llama:build-qwen-7b-cuda
  # For use without cuda `image: timothyswt/llama-cpp-server-amd64-qwen-vl-7b:latest`
  #
  # llama-vl-server-7b:
  #   image: timothyswt/llama-cpp-server-amd64-qwen-vl-7b-cuda:latest
  #   container_name: llama_server_vision_7b
  #   ports:
  #     - "8082:8080"  # Vision 7B API endpoint
  #   restart: unless-stopped
  #   healthcheck:
  #     test: ["CMD-SHELL", "curl -f http://localhost:8080/health || exit 1"]
  #     interval: 30s
  #     timeout: 10s
  #     retries: 3
  #     start_period: 90s  # Larger models take even longer
  #   networks:
  #     - mcp_network
  #   deploy:
  #     resources:
  #       limits:
  #         memory: 16G  # 7B models need more RAM
  #       reservations:
  #         memory: 8G
  #         devices:
  #           - driver: nvidia
  #             count: 1
  #             capabilities: [gpu]

  # mimir-server:
  #   build:
  #     context: .
  #     dockerfile: Dockerfile
  #     tags:
  #       - timothyswt/mimir:latest
  #   image: timothyswt/mimir:latest
  #   container_name: mimir_server
  #   environment:
  #     # Database Configuration
  #     - NEO4J_URI=bolt://nornicdb:7687
  #     - NEO4J_USER=admin
  #     - NEO4J_PASSWORD=${NEO4J_PASSWORD:-admin}
      
  #     # Server Configuration
  #     - NODE_ENV=production
  #     - PORT=3000
  #     # - NODE_TLS_REJECT_UNAUTHORIZED=${NODE_TLS_REJECT_UNAUTHORIZED:-1}
      
  #     # Workspace Configuration
  #     - WORKSPACE_ROOT=/workspace
  #     - HOST_WORKSPACE_ROOT=${HOST_WORKSPACE_ROOT}  # Pass through from host
  #     - HOST_HOME=${HOME}  # Host's home directory for expanding ~ in HOST_WORKSPACE_ROOT
     
  #     # LLM API Configuration
  #     - MIMIR_DEFAULT_PROVIDER=${MIMIR_DEFAULT_PROVIDER:-copilot}
  #     - MIMIR_LLM_API=${MIMIR_LLM_API:-http://copilot-api:4141}
  #     - MIMIR_LLM_API_PATH=${MIMIR_LLM_API_PATH:-/v1/chat/completions}
  #     - MIMIR_LLM_API_MODELS_PATH=${MIMIR_LLM_API_MODELS_PATH:-/v1/models}
  #     - MIMIR_LLM_API_KEY=${MIMIR_LLM_API_KEY:-dummy-key}
      
  #     # PCTX Integration (Code Mode for 90-98% token reduction)
  #     - PCTX_URL=${PCTX_URL:-http://host.docker.internal:8080}
  #     - PCTX_ENABLED=${PCTX_ENABLED:-false} 

      
  #     # Provider and Model Configuration (100% dynamic - no config file needed)
  #     - MIMIR_DEFAULT_MODEL=${MIMIR_DEFAULT_MODEL:-gpt-4.1}
      
  #     # Per-Agent Model Configuration (optional overrides)
  #     - MIMIR_PM_MODEL=${MIMIR_PM_MODEL:-}
  #     - MIMIR_WORKER_MODEL=${MIMIR_WORKER_MODEL:-}
  #     - MIMIR_QC_MODEL=${MIMIR_QC_MODEL:-}
      
  #     # Context Window Configuration
  #     - MIMIR_DEFAULT_CONTEXT_WINDOW=${MIMIR_DEFAULT_CONTEXT_WINDOW:-128000}
      
      
  #     # Embeddings API Configuration
  #     - MIMIR_EMBEDDINGS_PROVIDER=${MIMIR_EMBEDDINGS_PROVIDER:-openai}
  #     - MIMIR_EMBEDDINGS_API=${MIMIR_EMBEDDINGS_API:-http://llama-server:8080}
  #     - MIMIR_EMBEDDINGS_API_PATH=${MIMIR_EMBEDDINGS_API_PATH:-/v1/embeddings}
  #     - MIMIR_EMBEDDINGS_API_MODELS_PATH=${MIMIR_EMBEDDINGS_API_MODELS_PATH:-/v1/models}
  #     - MIMIR_EMBEDDINGS_API_KEY=${MIMIR_EMBEDDINGS_API_KEY:-dummy-key}
      
  #     # Embeddings Configuration
  #     - MIMIR_EMBEDDINGS_ENABLED=${MIMIR_EMBEDDINGS_ENABLED:-true}
  #     - MIMIR_EMBEDDINGS_MODEL=${MIMIR_EMBEDDINGS_MODEL:-bge-m3}
  #     - MIMIR_EMBEDDINGS_DIMENSIONS=${MIMIR_EMBEDDINGS_DIMENSIONS:-1024}
  #     - MIMIR_EMBEDDINGS_CHUNK_SIZE=${MIMIR_EMBEDDINGS_CHUNK_SIZE:-768}
  #     - MIMIR_EMBEDDINGS_CHUNK_OVERLAP=${MIMIR_EMBEDDINGS_CHUNK_OVERLAP:-100}
  #     - MIMIR_EMBEDDINGS_DELAY_MS=${MIMIR_EMBEDDINGS_DELAY_MS:-0}
  #     - MIMIR_EMBEDDINGS_MAX_RETRIES=${MIMIR_EMBEDDINGS_MAX_RETRIES:-3}
  #     # Image Embeddings Control (disabled by default for safety)
  #     - MIMIR_EMBEDDINGS_IMAGES=${MIMIR_EMBEDDINGS_IMAGES:-false}  # Default: disabled
  #     - MIMIR_EMBEDDINGS_IMAGES_DESCRIBE_MODE=${MIMIR_EMBEDDINGS_IMAGES_DESCRIBE_MODE:-true}  # Default: VL description mode
  #     # Qwen2.5-VL Configuration (for llama.cpp server)
  #     - MIMIR_EMBEDDINGS_VL_PROVIDER=${MIMIR_EMBEDDINGS_VL_PROVIDER:-llama.cpp}
  #     - MIMIR_EMBEDDINGS_VL_API=${MIMIR_EMBEDDINGS_VL_API:-http://llama-vl-server-2b:8080}
  #     - MIMIR_EMBEDDINGS_VL_API_PATH=${MIMIR_EMBEDDINGS_VL_API_PATH:-/v1/chat/completions}
  #     - MIMIR_EMBEDDINGS_VL_API_KEY=${MIMIR_EMBEDDINGS_VL_API_KEY:-dummy-key}
  #     - MIMIR_EMBEDDINGS_VL_MODEL=${MIMIR_EMBEDDINGS_VL_MODEL:-qwen2.5-vl}
  #     - MIMIR_EMBEDDINGS_VL_CONTEXT_SIZE=${MIMIR_EMBEDDINGS_VL_CONTEXT_SIZE:-131072}  # 128K tokens (7b/72b)
  #     - MIMIR_EMBEDDINGS_VL_MAX_TOKENS=${MIMIR_EMBEDDINGS_VL_MAX_TOKENS:-2048}  # Max description length
  #     - MIMIR_EMBEDDINGS_VL_TEMPERATURE=${MIMIR_EMBEDDINGS_VL_TEMPERATURE:-0.7}
  #     - MIMIR_EMBEDDINGS_VL_DIMENSIONS=${MIMIR_EMBEDDINGS_VL_DIMENSIONS:-768}  # Falls back to text model dims
  #     - MIMIR_EMBEDDINGS_VL_TIMEOUT=${MIMIR_EMBEDDINGS_VL_TIMEOUT:-180000}  # 3 minutes (VL processing is slow)
     
  #     # Indexing Configuration
  #     - MIMIR_INDEXING_THREADS=${MIMIR_INDEXING_THREADS:-1}
      
  #     # Feature Flags
  #     - MIMIR_FEATURE_PM_MODEL_SUGGESTIONS=${MIMIR_FEATURE_PM_MODEL_SUGGESTIONS:-true}
  #     - MIMIR_AUTO_INDEX_DOCS=${MIMIR_AUTO_INDEX_DOCS:-true}
      
  #     # Security Configuration
  #     - MIMIR_ENABLE_SECURITY=${MIMIR_ENABLE_SECURITY:-false}
  #     - MIMIR_DEV_USER_ADMIN=${MIMIR_DEV_USER_ADMIN}
  #     - MIMIR_DEV_USER_DEVELOPER=${MIMIR_DEV_USER_DEVELOPER}
  #     - MIMIR_DEV_USER_ANALYST=${MIMIR_DEV_USER_ANALYST}
  #     - MIMIR_DEV_USER_VIEWER=${MIMIR_DEV_USER_VIEWER}
  #     - MIMIR_JWT_SECRET=${MIMIR_JWT_SECRET}
  #     # OAuth Configuration (explicit endpoint URLs - provider-specific)
  #     - MIMIR_AUTH_PROVIDER=${MIMIR_AUTH_PROVIDER}
  #     - MIMIR_OAUTH_AUTHORIZATION_URL=${MIMIR_OAUTH_AUTHORIZATION_URL}  # Full authorization endpoint URL
  #     - MIMIR_OAUTH_TOKEN_URL=${MIMIR_OAUTH_TOKEN_URL}                  # Full token endpoint URL
  #     - MIMIR_OAUTH_USERINFO_URL=${MIMIR_OAUTH_USERINFO_URL}            # Full userinfo endpoint URL
  #     - MIMIR_OAUTH_CLIENT_ID=${MIMIR_OAUTH_CLIENT_ID}
  #     - MIMIR_OAUTH_CLIENT_SECRET=${MIMIR_OAUTH_CLIENT_SECRET}
  #     - MIMIR_OAUTH_CALLBACK_URL=${MIMIR_OAUTH_CALLBACK_URL}
  #     - MIMIR_OAUTH_ALLOW_HTTP=${MIMIR_OAUTH_ALLOW_HTTP}

  #     # Advanced Configuration
  #     - MIMIR_PARALLEL_EXECUTION=${MIMIR_PARALLEL_EXECUTION:-false}
  #     - MIMIR_INSTALL_DIR=${MIMIR_INSTALL_DIR:-/app}
  #     - MIMIR_AGENTS_DIR=${MIMIR_AGENTS_DIR:-/app/docs/agents}
  #   volumes:
  #     - ./data:/app/data
  #     - ./logs:/app/logs
  #     - ${HOST_WORKSPACE_ROOT:-~/src}:${WORKSPACE_ROOT:-/workspace}
  #   ports:
  #     - "9042:3000"
  #   healthcheck:
  #     test: ["CMD", "node", "-e", "require('http').get('http://localhost:3000/health', (res) => process.exit(res.statusCode === 200 ? 0 : 1)).on('error', () => process.exit(1))"]
  #     interval: 30s
  #     timeout: 10s
  #     retries: 3
  #     start_period: 10s
  #   depends_on:
  #     nornicdb:
  #       condition: service_healthy
  #     # copilot-api:
  #     #   condition: service_healthy
  #     # llama.cpp server for embeddings
  #     # llama-server:
  #     #   condition: service_healthy
  #     # llama.cpp server for image embeddings must set 
  #     # llama-vl-server:
  #       # condition: service_healthy
  #   networks:
  #     - mcp_network

networks:
  mcp_network:
    driver: bridge