docker-compose.yml

# ==============================================================================
# Thoth AI Research Assistant - Production Configuration
#
# IMPORTANT: Letta services are now INDEPENDENT!
# Start Letta first: docker compose -f docker-compose.letta.yml up -d
# Then start Thoth: docker compose up -d
#
# MICROSERVICES ARCHITECTURE:
#    - Runs each service in separate containers
#    - 5 containers: api, mcp, monitor, dashboard, discovery
#    - Connects to external Letta services
#    - Resource usage: ~4GB RAM
#
# ==============================================================================

# ==============================================================================
# Shared Configuration
# ==============================================================================
x-thoth-common: &thoth-common
  # NOTE: Individual services override build config with service-specific Dockerfiles
  # This common config is for shared runtime settings only
  user: "${USER_ID:-1000}:${GROUP_ID:-1000}"
  environment: &thoth-env
    # Database connections (connects to external Letta postgres)
    - DATABASE_URL=postgresql://thoth:thoth_password@letta-postgres:5432/thoth

    # Vault-relative path configuration
    - OBSIDIAN_VAULT_PATH=/vault
    - THOTH_SETTINGS_FILE=/vault/thoth/_thoth/settings.json

    # Operational settings
    - THOTH_LOG_LEVEL=INFO
    - THOTH_DOCKER=1
    - DOCKER_ENV=true
    - PYTHONPATH=/app/src

    # Fix Letta filesystem permissions
    - LETTA_LOG_FILE=/vault/thoth/_thoth/logs/letta.log
    - HOME=/vault/thoth/_thoth
    - THOTH_LOG_FILE=/vault/thoth/_thoth/logs/thoth.log

    # API Keys
    - API_MISTRAL_KEY=${API_MISTRAL_KEY:-}
    - API_OPENROUTER_KEY=${API_OPENROUTER_KEY:-}
    - API_OPENCITATIONS_KEY=${API_OPENCITATIONS_KEY:-}
    - API_GOOGLE_KEY=${API_GOOGLE_KEY:-}
    - API_SEMANTIC_SCHOLAR_KEY=${API_SEMANTIC_SCHOLAR_KEY:-}
    - API_WEB_SEARCH_KEY=${API_WEB_SEARCH_KEY:-}

    # Playwright browser path (fixed location, independent of HOME)
    - PLAYWRIGHT_BROWSERS_PATH=/app/.cache/ms-playwright

    # Discovery settings
    - THOTH_DISCOVERY_AUTO_START_SCHEDULER=false
    - THOTH_DISCOVERY_DEFAULT_MAX_ARTICLES=50

    # Agent settings
    - THOTH_RESEARCH_AGENT_MEMORY_ENABLED=true
    - THOTH_AGENT_MAX_TOOL_CALLS=20
    - THOTH_AGENT_TIMEOUT_SECONDS=300

    # Multi-user mode (set THOTH_MULTI_USER=true to enable)
    # When enabled, each user gets an isolated vault under THOTH_VAULTS_ROOT
    # and all data is scoped by user_id in the database.
    - THOTH_MULTI_USER=${THOTH_MULTI_USER:-false}
    - THOTH_VAULTS_ROOT=${THOTH_VAULTS_ROOT:-/vaults}
    - THOTH_ALLOW_REGISTRATION=${THOTH_ALLOW_REGISTRATION:-false}

  volumes: &thoth-volumes
    - ${OBSIDIAN_VAULT_PATH:?Please set OBSIDIAN_VAULT_PATH to your Obsidian vault root directory}:/vault:rw
    - thoth-tmp-cache:/tmp/cache
    - ./templates:/app/templates:ro
    - thoth-app-home:/home/thoth
    # Multi-user vault root: mounted only when THOTH_MULTI_USER=true.
    # Set THOTH_VAULTS_ROOT on the host and it will be bind-mounted at /vaults.
    # In single-user mode this mount is harmless (unused directory).
    - ${THOTH_VAULTS_ROOT:-./vaults}:/vaults:rw

  networks:
    - thoth-network
    - letta-network  # Connect to external Letta network

  restart: unless-stopped

services:
  # ==============================================================================
  # Thoth API Server
  # Main REST API for frontend and external integrations
  # ==============================================================================
  thoth-api:
    <<: *thoth-common
    image: ghcr.io/acertainknight/project-thoth/api:latest
    build:
      context: .
      dockerfile: docker/api/Dockerfile
      target: production
    pull_policy: always
    container_name: thoth-api
    command: ["python", "-m", "thoth", "server", "start", "--api-host", "0.0.0.0", "--api-port", "8000"]
    ports:
      - "8080:8000"
    # NO DEPENDENCIES - connects to external Letta
    healthcheck:
      test: ["CMD", "python", "-c", "import requests; requests.get('http://localhost:8000/health', timeout=5)"]
      interval: 30s
      timeout: 10s
      retries: 5
      start_period: 60s
    deploy:
      resources:
        limits:
          memory: 1G
          cpus: '0.5'
        reservations:
          memory: 256M
          cpus: '0.1'

  # ==============================================================================
  # Thoth MCP Server
  # Model Context Protocol server for Claude integration
  # ==============================================================================
  thoth-mcp:
    <<: *thoth-common
    image: ghcr.io/acertainknight/project-thoth/mcp:latest
    build:
      context: .
      dockerfile: docker/mcp/Dockerfile
      target: production
    pull_policy: always
    container_name: thoth-mcp
    # HTTP-only mode (includes /mcp POST and /sse streaming endpoints)
    command: ["python", "-m", "thoth", "mcp", "http", "--host", "0.0.0.0", "--port", "8000"]
    # NO DEPENDENCIES - connects to external Letta
    environment:
      # Base Thoth environment
      - DATABASE_URL=postgresql://thoth:thoth_password@letta-postgres:5432/thoth
      - OBSIDIAN_VAULT_PATH=/vault
      - THOTH_SETTINGS_FILE=/vault/thoth/_thoth/settings.json
      - THOTH_LOG_LEVEL=INFO
      - THOTH_DOCKER=1
      - DOCKER_ENV=true
      - PYTHONPATH=/app/src
      - LETTA_LOG_FILE=/vault/thoth/_thoth/logs/letta.log
      - HOME=/vault/thoth/_thoth
      - THOTH_LOG_FILE=/vault/thoth/_thoth/logs/thoth.log
      # MCP-specific settings
      - MCP_HOST=0.0.0.0
      - MCP_HTTP_PORT=8000
      - MCP_ENABLE_FILES=true
    ports:
      - "8082:8000"  # HTTP transport (includes /mcp and /sse endpoints)
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:8000/health"]
      interval: 30s
      timeout: 10s
      retries: 5
      start_period: 60s
    # networks inherited from thoth-common (thoth-network + letta-network)
    deploy:
      resources:
        limits:
          memory: 512M
          cpus: '0.25'
        reservations:
          memory: 128M
          cpus: '0.05'

  # ==============================================================================
  # Thoth PDF Monitor
  # Watches vault directory for new PDFs and processes them automatically
  # ==============================================================================
  thoth-monitor:
    <<: *thoth-common
    image: ghcr.io/acertainknight/project-thoth/pdf-monitor:latest
    build:
      context: .
      dockerfile: docker/pdf-monitor/Dockerfile
      target: production
    pull_policy: always
    container_name: thoth-monitor
    command: ["python", "-m", "thoth", "monitor", "--optimized", "--recursive", "--watch-dir", "/vault/thoth/papers"]
    depends_on:
      thoth-api:
        condition: service_started
    healthcheck:
      test: ["CMD-SHELL", "kill -0 1"]
      interval: 60s
      timeout: 10s
      retries: 3
      start_period: 30s
    deploy:
      resources:
        limits:
          memory: 2G
          cpus: '1.5'
        reservations:
          memory: 512M
          cpus: '0.25'

  # ==============================================================================
  # Thoth Discovery Dashboard
  # Automatically exports new discovery results and imports sentiment changes
  # ==============================================================================
  thoth-dashboard:
    <<: *thoth-common
    build:
      context: .
      dockerfile: Dockerfile
      args:
        SERVICE_EXTRAS: "service-dashboard"
    container_name: thoth-dashboard
    command: [
      "python", "-m", "thoth", "research", "start-dashboard",
      "--check-interval", "60"
    ]
    depends_on:
      thoth-api:
        condition: service_started
    healthcheck:
      test: [
        "CMD", "python", "-c",
        "from pathlib import Path; import sys; sys.exit(0 if Path('/vault/Research/Dashboard').exists() else 1)"
      ]
      interval: 60s
      timeout: 10s
      retries: 3
      start_period: 30s
    deploy:
      resources:
        limits:
          memory: 512M
          cpus: '0.25'
        reservations:
          memory: 128M
          cpus: '0.05'

  # ==============================================================================
  # Thoth Discovery Scheduler
  # Runs automated daily paper discovery searches for research questions
  # ==============================================================================
  thoth-discovery:
    <<: *thoth-common
    image: ghcr.io/acertainknight/project-thoth/discovery:latest
    build:
      context: .
      dockerfile: docker/discovery/Dockerfile
      target: production
    pull_policy: always
    container_name: thoth-discovery
    command: [
      "python", "-m", "thoth", "research", "scheduler"
    ]
    depends_on:
      thoth-api:
        condition: service_started
    healthcheck:
      test: [
        "CMD", "python", "-c",
        "from thoth.services.discovery_scheduler import ResearchQuestionScheduler; import sys; sys.exit(0)"
      ]
      interval: 60s
      timeout: 10s
      retries: 3
      start_period: 30s
    deploy:
      resources:
        limits:
          memory: 512M
          cpus: '0.25'
        reservations:
          memory: 128M
          cpus: '0.05'

# ==============================================================================
# Persistent Volumes
# ==============================================================================
volumes:
  # Thoth-specific volumes
  thoth-app-home:
    driver: local
    name: thoth-app-home

  thoth-tmp-cache:
    driver: local
    name: thoth-tmp-cache

# ==============================================================================
# Networks
# ==============================================================================
networks:
  thoth-network:
    driver: bridge
    name: thoth-network
    ipam:
      config:
        - subnet: 172.20.0.0/16

  # External Letta network - must be created by docker-compose.letta.yml
  letta-network:
    external: true
    name: letta-network