LabscoMCP early access ↗
RajawatrajatMCP SERVER

MCP Software Engineer Server

Scaffold a full-stack project and keep going — database migrations, API endpoints, auth, components, Docker, CI/CD and tests, each as its own tool.

by Rajawatrajat ↗Code Generation, Scaffolding & MigrationCommunity

LABSCO SUMMARY

Wide surface, and the write tools are real writes.

The value is that setup decisions become named calls with named arguments — "Prisma against PostgreSQL at this path" instead of a paragraph of instructions the model interprets. The cost is scope: this server creates directories, runs migrations, commits to git and deploys to clouds, all through an assistant. Point it at a scratch project first, and check what deploy_to_cloud and backup_database are configured against before you let either run.

What it is

A broad MCP server that turns the repetitive parts of full-stack work into tool calls. Rather than asking a model to write a Dockerfile from memory, it calls a tool that generates one for the framework you named. The coverage runs from project creation through database, frontend, backend, deployment, testing and git.

What you get

  • A project initialised on a chosen framework with the feature set you name, plus file operations to work inside it — create_project, read_file, write_file, create_directory, list_files, search_files
  • A database wired up through an ORM, with migrations, generated models, seeds, backups and direct queries — init_database, create_migration, run_migrations, generate_model, seed_database, backup_database, query_database
  • Frontend scaffolding: components, pages, routing, state management, form handling with validation, a CSS framework configured, and bundle optimisation — create_component, create_page, setup_routing, setup_state_management, setup_forms, setup_styling, optimize_bundle
  • Backend scaffolding: REST endpoints with CRUD, an authentication scheme, middleware, request validation, WebSockets, background jobs and logging — create_api_endpoint, setup_authentication, create_middleware, setup_validation, setup_websockets, create_background_job, setup_logging
  • Deployment artefacts: Dockerfiles, Docker Compose, a cloud deploy, a CI/CD pipeline, SSL, a load balancer and monitoring — create_dockerfile, create_docker_compose, deploy_to_cloud, setup_ci_cd, setup_ssl, setup_load_balancer, setup_monitoring
  • Testing and version control: a test framework configured, test files generated, tests run with coverage, a security scan, and repository, branch, commit and hook operations — setup_testing, create_test, run_tests, security_scan, init_repository, create_branch, commit_changes, setup_hooks

Requirements

Node.js 18 or higher, npm or yarn, and git. Docker if you want the containerisation tools to do anything, plus whichever language runtime the project type needs. There is no published package: clone, npm install, npm run build, and point your client at dist/index.js by absolute path. It writes to your filesystem and can deploy to cloud providers, so give it a project path you are willing to have changed.

Setup effort

Build from source — git clone https://github.com/Rajawatrajat/mcp-software-engineer.git

See installation options