DevFlow MCP Server
About
DevFlow MCP Server is a professional Model Context Protocol (MCP) server for development‑workflow automation. It lets you create and run language‑agnostic tools through an extensible plugin system supporting C#, TypeScript, and Python, using Clean Architecture to separate…
Details
- Author
- maxxentropy
- Downloads
- 136
- Categories
- Other
Jump to
- MCP 1.0 compliant REST/JSON‑RPC façade
- Hot‑loadable plugins in C#, TypeScript, and Python
- Workflow engine (create, run, query, list)
- Built‑in health check endpoint
- PowerShell test scripts for CI/CD smoke tests
- Clean Architecture with CQRS and Mediator pattern
Ensure .NET SDK 8.0 and PowerShell 7.x are installed. Build with dotnet build in the src directory, then run dotnet run from DevFlow.Host. The server listens on http://localhost:5000. Use the test script .\Test-Simple.ps1 to verify health, list plugins, and execute the HelloWorldPlugin.
DevFlow MCP Server
A professional Model Context Protocol (MCP) server for development-workflow automation.
DevFlow lets you create and run language-agnostic tools through a clean, extensible plugin system supporting C#, TypeScript, and Python.
Its layered architecture (Clean Architecture) keeps business logic, infrastructure, and presentation concerns neatly separated.
---
Table of Contents
1. Features
2. Quick Start
3. Available Tools
4. Testing the Setup
5. Architecture Overview
6. Plugin System
7. Troubleshooting
8. Manual API Testing
9. License
---
Features
- MCP 1.0 compliant REST/JSON-RPC façade
- Hot-loadable plugins in C#, TypeScript, and Python
- Workflow engine (create, run, query, list)
- Built-in health check endpoint
- PowerShell test scripts for CI/CD smoke tests
- Clean Architecture with CQRS & Mediator pattern
---
Quick Start
Prerequisites
| Software | Version |
|------------|---------|
| .NET SDK | 8.0 |
| PowerShell | 7.x |
1 — Build the solution
cd src
dotnet build
2 — Run the server
cd DevFlow.Host
dotnet run # → http://localhost:5000
3 — Smoke‑test the install
```powershell
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.



