mcpField
M
MCPStudio: The Postman for Model Context Protocol

MCPStudio: The Postman for Model Context Protocol

Local
github.com/lonexrebDeveloper Tools2Updated 30w ago

About MCPStudio: The Postman for Model Context Protocol

MCPStudio is a web platform for creating, testing, managing, and discovering MCP servers, aimed at developers who want a visual interface for working with them. It also supports execution comparisons, visual pipelines, and execution history analysis.

Key features

  1. 01Create, connect, and monitor MCP servers
  2. 02Discover and run tools with result views
  3. 03Search tools and resources across servers
  4. 04Build DAG pipelines from tool calls
  5. 05Compare runs and inspect JSON differences
  6. 06Search npm and GitHub server registries

Use cases

  1. 01Check server status and find available tools
  2. 02Test tool parameters and inspect results
  3. 03Chain tools into a DAG workflow
  4. 04Compare tool outputs and execution times

MCP support

MCPStudio acts as an MCP client for connecting to and managing servers, discovering and executing their tools, browsing resources, and checking server status. It can also search MCP server listings on npm and GitHub.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

No install method listed yet

We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.

Repository

Tools

No tool list yet

Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.

Details

From the project README

MCPStudio Banner MCPStudio Banner

MCPStudio: The Postman for Model Context Protocol

MCPStudio is a web platform for creating, testing, managing, and discovering Model Context Protocol (MCP) servers. Like Postman for APIs, MCPStudio provides a visual interface that abstracts away protocol complexity — letting you connect to MCP servers, discover tools, execute them with parameters, and inspect results, all from the browser.

Strategic direction: Web app (now) → CLI tool → SDK/library → Desktop app

Features

  • Server Management — Create, configure, connect to, and monitor MCP servers with status tracking (connected / disconnected / error)
  • Tool Discovery & Execution — Auto-discover tools exposed by MCP servers, fill parameters via dynamic forms, and view formatted responses
  • Tools Library — Browse all tools across connected servers with search and server filtering
  • Execution Dashboard — Real-time metrics with Recharts (timing, tool distribution, success/error rates) and persistent history via IndexedDB
  • Execution History — Server-side persistent execution log with expandable rows, pagination, and clear
  • Visual Pipeline Builder — Drag-and-drop React Flow canvas to chain MCP tools into DAG pipelines with auto-layout and frontend execution engine
  • Tool Execution Arena — Side-by-side resizable panels for comparing tool executions with JSON diff and timing comparison
  • MCP Server Discovery — Search npm and GitHub registries for MCP-compatible servers with source filtering and install commands
  • Resources Browser — Discover and browse data sources exposed by connected MCP servers
  • Prompt Templates — Create, edit, tag, and manage reusable prompt patterns with {{variable}} placeholders (stored in IndexedDB)
  • Config Export/Import — Export server configs as JSON or YAML (credentials redacted); import via file upload, drag & drop, or paste
  • Authentication — Supabase auth (email/password signup & login) with JWT tokens; OAuth2 flow support for service integrations (Google Drive)
  • Guided Onboarding — First-run splash screen with interactive spotlight tour and confetti celebration
  • Settings — Theme toggle (dark/light), API base URL configuration, profile, data management
  • Documentation & Support — Built-in docs, FAQ, quick links, and feedback form
  • Real-time Updates — WebSocket streams for tool execution and server status changes
  • Google Drive Integration — Connect via OAuth, list/search/retrieve files, create folders

Tech Stack

Frontend
CategoryTechnology
FrameworkReact 18 + TypeScript
BuildVite 5 + SWC
StateZustand 5 (persisted to localStorage)
Data FetchingReact Query 5
UI Kitshadcn/ui (47+ Radix-based components)
StylingTailwind CSS 3 with custom MCP color tokens + gradients
Formsreact-hook-form + Zod validation
Routingreact-router-dom 6
ChartsRecharts 2
Node Graph@xyflow/react (React Flow) + @dagrejs/dagre
Auth@supabase/supabase-js (email/password)
Local DBDexie (IndexedDB wrapper) — executions, pipelines, prompts
Config Parsingjs-yaml + Zod
Effectscanvas-confetti (onboarding celebration)
NotificationsSonner
Backend
CategoryTechnology
FrameworkFastAPI 0.109
LanguagePython 3.10+
DatabaseMongoDB (Motor async driver) — falls back to in-memory mock
AuthJWT (python-jose) + OAuth2 (google-auth-oauthlib)
DIdependency-injector
HTTP Clienthttpx (async, for registry search)
WebSocketswebsockets 12
ValidationPydantic 2 + pydantic-settings
Testingpytest + pytest-asyncio

Architecture

Backend — Domain-Driven Design (4 layers)
backend/mcp_studio_backend/src/mcp_studio/
├── api/              # Controllers, routes, Pydantic schemas
│   ├── controllers/  # auth, server, tool, execution, discovery
│   ├── routes/       # auth, server, tool, execution, discovery
│   └── schemas/      # auth, server, tool, execution, discovery
├── application/      # Services, DTOs (orchestration)
│   └── services/     # auth, server, tool, execution, discovery
├── domain/           # Models, repository interfaces, domain services
├── infrastructure/   # MongoDB repos, Google Drive client, registry clients, EventBus
├── config/           # Settings (pydantic-settings, loads .env)
├── container.py      # Dependency injection wiring
└── main.py           # FastAPI app entry point
Frontend — Feature-Sliced Architecture
frontend/ai-server-forge/src/
├── features/
│   ├── auth/           # Login, Signup, AuthGuard, auth store, Supabase hooks
│   ├── servers/        # Dashboard, ServerDetail, ConfigExport/Import, use-servers
│   ├── tools/          # ToolEditor, ParameterEditor, use-tools
│   ├── tools-library/  # ToolsLibrary page (cross-server tool browser)
│   ├── pipeline/       # PipelineList, PipelineEditor, React Flow canvas, engine
│   ├── execution/      # Arena, ExecutionPanel, ExecutionHistory, ExecutionHistoryPage, Metrics
│   ├── resources/      # ResourcesPage, ResourceCard, use-resources
│   ├── prompts/        # PromptsPage, PromptCard, PromptEditor (Dexie/IndexedDB)
│   ├── discovery/      # DiscoveryPage, DiscoveryCard (npm + GitHub search)
│   ├── settings/       # SettingsPage, ThemeToggle, ApiConfig, DataManagement
│   ├── onboarding/     # SplashScreen, TourOverlay (spotlight + confetti)
│   ├── docs/           # DocsPage (static documentation)
│   └── support/        # SupportPage, FAQ, QuickLinks, FeedbackForm
├── components/
│   ├── layout/       # MainLayout, Header, Sidebar
│   └── ui/           # shadcn/ui component library (47+ components)
├── stores/           # Zustand stores (ui)
├── lib/
│   ├── api/          # Typed API client (servers, tools, auth, resources, executions, discovery)
│   └── supabase.ts   # Supabase client initialization
├── hooks/            # Shared hooks (use-mobile, use-toast)
└── types/            # TypeScript types matching backend Pydantic schemas

Getting Started

Prerequisites
  • Python 3.10+
  • Node.js 16+
  • MongoDB (optional — falls back to in-memory mock when unavailable)
  • UV package manager (recommended)
Installation
git clone https://github.com/lonexreb/MCPstudio.git
cd MCPstudio

Backend:

cd backend/mcp_studio_backend
python -m venv venv && source venv/bin/activate  # Windows: venv\Scripts\activate
pip install uv && uv pip install -e .
uvicorn mcp_studio.main:app --reload  # runs on :8000

Frontend:

cd frontend/ai-server-forge
npm install
npm run dev  # runs on :8080

Open your browser at http://localhost:8080 and sign up for a new account, or log in with existing credentials.

API Endpoints

MethodPathPurpose
POST/api/auth/registerRegister new user
POST/api/auth/tokenLogin (OAuth2 password flow)
GET/api/auth/meCurrent user info
GET/api/auth/google/authGoogle OAuth URL
GET/api/auth/google/callbackGoogle OAuth callback
POST/api/serversCreate server
GET/api/serversList servers
GET/api/servers/{id}Get server details
PUT/api/servers/{id}Update server
DELETE/api/servers/{id}Delete server
POST/api/servers/{id}/connectConnect & discover tools
POST/api/servers/{id}/disconnectDisconnect
GET/api/servers/{id}/toolsList server tools
GET/api/servers/{id}/resourcesList server resources
GET/api/toolsList all tools (cross-server)
GET/api/tools/{id}Get tool details
POST/api/servers/{id}/tools/{id}/executeExecute tool
GET/api/executionsExecution history (filterable)
GET/api/executions/{id}Get execution details
DELETE/api/executionsClear execution history
GET/api/discovery/searchSearch MCP servers (npm/GitHub)
GET/api/discovery/categoriesList server categories
WS/ws/servers/{id}/tools/{id}/executionTool execution stream

Roadmap

MCPStudio follows a phased enhancement plan inspired by patterns from Unsloth Studio:

PhaseFeatureStatus
0Foundation — Zustand stores, React Query, typed API client, auth guardDone
1Feature-sliced architecture reorganizationDone
2UX — Server Connection Wizard, terminal-style loading, collapsible sectionsDone
3Real-time Execution Dashboard — Recharts metrics, IndexedDB history (Dexie)Done
4Visual MCP Pipeline Builder — React Flow canvas, DAG execution, dagre auto-layoutDone
5Tool Execution Arena — side-by-side comparison with JSON diffDone
6Config Export/Import — JSON/YAML with Zod validation, drag & drop importDone
7Supabase Auth — email/password signup & login, replacing hardcoded credentialsDone
8Fill the Gaps — Settings, Tools Library, Resources, Prompts, Docs, Support pagesDone
9Onboarding & Guided Tour — splash screen, spotlight tour, confetti celebrationDone
10Execution History — server-side persistence, history page with paginationDone
11MCP Server Discovery — npm/GitHub registry search with discovery pageDone

See EXPERIMENT.md for detailed tracking.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License — see the LICENSE file for details.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.