Skip to main content
Glama
docker-compose.yml2.24 kB
services: spec-workflow-mcp: build: context: .. dockerfile: containers/Dockerfile ports: # Security: Only expose to host's localhost by default # To expose to network: DASHBOARD_HOST=0.0.0.0 docker-compose up - "${DASHBOARD_HOST:-127.0.0.1}:${DASHBOARD_PORT:-5000}:${DASHBOARD_PORT:-5000}" volumes: # Global state directory - shared between host MCP clients and container # Uses SPEC_WORKFLOW_HOME inside container for state files (activeProjects.json, session, etc.) - "${HOME}/.spec-workflow-mcp:/home/node/.spec-workflow-mcp:rw" # Mount host projects directory for path translation # Set HOST_PROJECT_ROOT to your projects parent directory (e.g., /Users/yourname or /home/yourname) - "${HOST_PROJECT_ROOT:-${HOME}}:/projects:rw" environment: - DASHBOARD_PORT=${DASHBOARD_PORT:-5000} # Global state location (upstream's SPEC_WORKFLOW_HOME) # Required for sandboxed environments where container $HOME may differ - SPEC_WORKFLOW_HOME=/home/node/.spec-workflow-mcp # Path translation for Docker: maps host paths to container paths # When host MCP client registers /Users/dev/myproject, container accesses /projects/myproject - SPEC_WORKFLOW_HOST_PATH_PREFIX=${HOST_PROJECT_ROOT:-${HOME}} - SPEC_WORKFLOW_CONTAINER_PATH_PREFIX=/projects # Docker networking: App binds to 0.0.0.0 inside container (required for port forwarding) # External access is controlled by Docker port mapping above, not these settings - SPEC_WORKFLOW_BIND_ADDRESS=0.0.0.0 - SPEC_WORKFLOW_ALLOW_EXTERNAL_ACCESS=true # Security: Rate limiting (defaults to enabled) - SPEC_WORKFLOW_RATE_LIMIT_ENABLED=${SPEC_WORKFLOW_RATE_LIMIT_ENABLED:-true} restart: unless-stopped # Security: Read-only root filesystem (except mounted volumes) read_only: true # Security: Drop all capabilities cap_drop: - ALL # Security: No new privileges security_opt: - no-new-privileges:true # Resource limits to prevent DoS deploy: resources: limits: cpus: '1.0' memory: 512M reservations: cpus: '0.25' memory: 128M

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/Pimzino/spec-workflow-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server