Skip to main content
Glama

Multi Agent Orchestrator MCP

An enterprise‑grade Model Context Protocol (MCP) server for autonomous software engineering. It coordinates specialized agents (Architecture, Quality, Cloud, Prompt) to plan, build, test, and deploy applications with self‑healing, authentication, and analytics.

Check out!

Smithery Platfrom Deployed Link: https://smithery.ai/server/@yoriichi-07/multi_orchestrator_mcp

Team

  • Team Name : UpsideDown

  • Member : Shreesaanth R

Hackathon

  • Theme 2: Build a Secure MCP Server for Agents (w/ Cequence)

  • Challenge addressed: Build a production‑ready MCP server that orchestrates multiple agents with authentication (Descope), hosting (Cequence), and self‑healing to reliably execute end‑to‑end development workflows, deployable on Smithery.


Requirements

  • Python 3.11+

  • Git

  • An MCP‑compatible client (VS Code, Cursor, Windsurf, Claude Desktop, etc.)


Getting started

First, install the server with your MCP client. For an overview of client support and mechanics, see the official MCP quickstart: https://modelcontextprotocol.io/quickstart/user

Standard config (works in most clients)

{ "mcpServers": { "multi_orchestrator_mcp": { "command": "cmd", "args": [ "/c", "npx", "-y", "@smithery/cli@latest", "run", "@yoriichi-07/multi_orchestrator_mcp", "--key", "70fd8cf1-9dd3-4556-8a43-78916f617fb2" ] } } }
  • One Click installation

# VS Code (stable) npx -y @smithery/cli@latest install @yoriichi-07/multi_orchestrator_mcp --client vscode --key 70fd8cf1-9dd3-4556-8a43-78916f617fb2 # VS Code Insiders npx -y @smithery/cli@latest install @yoriichi-07/multi_orchestrator_mcp --client vscode-insiders --key 70fd8cf1-9dd3-4556-8a43-78916f617fb2

Click the button to install (if the deeplink is not supported on your OS, use the manual steps below):

Manual: Go to Cursor SettingsMCPAdd new MCP Server. Choose command type and set:

npx -y @smithery/cli@latest install @yoriichi-07/multi_orchestrator_mcp --client cursor --key 70fd8cf1-9dd3-4556-8a43-78916f617fb2

Use the CLI then paste the standard config above:

claude mcp add --transport http yoriichi-07-multi-orchestrator-mcp "https://server.smithery.ai/@yoriichi-07/multi_orchestrator_mcp/mcp"

Follow the MCP quickstart: https://modelcontextprotocol.io/quickstart/user. Use the standard config above.

Use the Install → Edit mcp.json, paste the standard config. One-click: Add MCP

Docs: https://docs.windsurf.com/windsurf/cascade/mcp. Use the standard config above (replace URL with http://localhost:8080 for local).

# Clone and install git clone https://github.com/yoriichi-07/Multi_Orchestrator_MCP.git cd Multi_Orchestrator_MCP pip install -r requirements.txt # Configure (optional) cp config/env.template .env # edit .env with DESCOPE_* if you want auth and analytics # Start the MCP server (HTTP transport) python mcp_server.py # Default: http://localhost:8080 # MCP discovery path is served by FastMCP under the mounted app
docker build -t multi-orchestrator-mcp . docker run --rm -p 8080:8080 --env-file .env multi-orchestrator-mcp

Configuration

Environment variables (see config/env.template):

  • DESCOPE_PROJECT_ID, DESCOPE_MANAGEMENT_KEY, DESCOPE_ACCESS_KEY – enable Descope authentication (optional for local/dev).

  • PORT – server port (default 8080).

  • DESCOPE_DEMO_MODE – set true for local testing without full auth.

  • CEQUENCE_GATEWAY_ID, CEQUENCE_API_KEY – enable Cequence analytics (optional).

  • JWT_SECRET_KEY, CORS_ORIGINS, RATE_LIMIT_REQUESTS, logging toggles.

Client configuration templates are provided in config/mcp.json.template (direct JWT or auto‑refresh proxy modes).


Capabilities

  • orchestrate_task

    • Title: Orchestrate multi‑agent task

    • Description: Coordinate Frontend/Backend/DevOps/QA agents for development, testing, or deployment.

    • Parameters: task_description (string), task_type (enum: development|architecture|testing|deployment), priority (enum)

  • generate_architecture

    • Title: Generate architecture

    • Description: Produce system architecture with components and recommendations.

    • Parameters: project_description (string), tech_stack (string[]), requirements (string[])

  • auto_fix_code

    • Title: Self‑healing fix

    • Description: Generate fixes for code using error context and explanations.

    • Parameters: code (string), error_message (string), context (string)

  • list_capabilities

    • Title: Catalog

    • Description: Summarize available agents, tools, enterprise features, and supported tasks.

    • Parameters: none

  • get_system_status

    • Title: System status

    • Description: Returns server health, agent availability, analytics/auth status, and timestamp.

    • Parameters: none

  • advanced_generate_application

    • Title: Enterprise app generation

    • Description: Plan and generate an application using advanced agents and deployment strategies.

    • Parameters: description (string), complexity_level (enum), innovation_requirements (string[]), deployment_strategy (enum)

  • autonomous_architect

    • Title: Autonomous architect

    • Description: Builds an execution DAG and adaptive strategy from goals and constraints.

    • Parameters: project_goals (string[]), constraints (string[]), learning_objectives (string[])

  • proactive_quality_assurance

    • Title: Proactive quality

    • Description: Applies policy‑as‑code checks with optional auto‑remediation.

    • Parameters: code_context (string), quality_standards (string[]), auto_remediation (bool)

  • evolutionary_prompt_optimization

    • Title: Prompt evolution

    • Description: Creates and evolves prompts based on goals and performance metrics.

    • Parameters: base_prompt (string), optimization_goals (string[]), performance_metrics (object)

  • last_mile_cloud_deployment

    • Title: Cloud deployment

    • Description: Plans deployment, verifies environments, and returns rollback/monitoring setup.

    • Parameters: application_context (string), target_environments (string[]), verification_requirements (string[])

  • ping

    • Title: Health check

    • Description: Simple liveness probe.

    • Parameters: none

  • debug_server_config

    • Title: Debug configuration (temporary)

    • Description: Exposes non‑secret configuration metadata for diagnostics.

    • Parameters: none

  • mcp://capabilities — capabilities and catalog (JSON)

  • mcp://analytics — analytics snapshot (requires Cequence)

  • mcp://health — system health snapshot

  • project-setup — guided setup plan

  • code-review — structured review outline

  • revolutionary-development — advanced strategy plan using autonomous agents


Tech stack

Full dependencies are declared in requirements.txt and pyproject.toml.


Demo video

Link: (coming soon)


Future Roadmap

Enhanced Agent Intelligence

  • Implement reinforcement learning for agent coordination optimization

  • Add context‑aware agents that learn from project history and patterns

  • Develop specialized agents for mobile, ML, and blockchain development

Advanced Automation

  • Build predictive analytics to forecast development challenges

  • Create automated testing strategies with comprehensive edge case generation

  • Implement intelligent resource optimization for cloud deployments

Enterprise Features

  • Add multi‑tenant architecture with organization‑specific agent training

  • Implement advanced compliance standards (SOC2, HIPAA, PCI‑DSS)

  • Create custom agent marketplace for domain‑specific development patterns

Developer Experience

  • Build a visual development interface for non‑technical users

  • Integrate with Jira, GitHub Actions, Slack, and incident tooling

  • Enable real‑time collaboration for distributed development teams


-
security - not tested
F
license - not found
-
quality - not tested

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/yoriichi-07/Multi_Orchestrator_MCP'

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