Explore MCP Servers
Like a Pro
Browse tools, run prompts, inspect resources, and chat with LLMs — all over live MCP connections. The complete developer workbench for the Model Context Protocol.
10 Killer Features
Everything a developer needs to explore, debug, and build on top of MCP servers.
Streaming Chat with Auto Tool Calling
Real-time SSE-streamed AI conversations where the LLM automatically invokes MCP tools mid-response. Tool inputs, outputs, and token usage are shown inline — no black boxes.
Learn more →Workflow Builder with Load Testing
Chain MCP tools into multi-step pipelines, pass outputs between steps, track every run in history, and stress-test your server with concurrent load tests — all from the browser.
Learn more →Sensitive Data Protection
Heuristic + regex detection of secrets, API keys, JWTs, and passwords in tool parameters and chat messages. AES-256 encryption at rest. Configurable masking rules. Show/hide toggle on demand.
Learn more →Multi-Transport MCP Connections
Connect to MCP servers over stdio, SSE, or HTTP with custom headers, OAuth credentials, and Azure auth. Import/export encrypted connection sets to share with your team.
Learn more →10 Persisted Themes
Command Dark/Light, Nord, Dracula, Catppuccin Mocha, Solarized Light, GitHub Dark/Light, Material Dark/Light — each persisted per user. Switch with zero friction.
Learn more →Command Palette + Slash Commands
⌘K / Ctrl+K command palette for instant keyboard-first navigation. Slash commands in chat for quick actions — /prompt, /clear, /help. Designed for developer speed.
Learn more →Multi-Provider LLM Support
OpenAI, Azure OpenAI, Azure AI Foundry, Ollama, LM Studio, and any OpenAI-compatible endpoint. API keys are masked with show/hide toggle. Switch models per conversation.
Learn more →Dynamic Tool Parameter Forms
Tool parameter forms are generated from JSON Schema automatically — strings, numbers, booleans, enums (radio or dropdown), objects, and arrays all have native controls. No manual form building.
Learn more →Prompts with LLM Pipeline
Execute MCP prompts with dynamic arguments and pipe results directly into Chat — enabling prompt → LLM analysis pipelines in two clicks. Favourites persist across sessions.
Learn more →Single-Container, Zero Dependencies
One Docker image, one volume mount, done. No database, no external services. Also supports docker compose for separate frontend/backend deployment. Runs anywhere Docker runs.
Learn more →Quick Start
Up and running in under 2 minutes.
Pull & Run
Single container — no compose file needed.
docker run -d \
-p 8090:8080 \
-v mcp-data:/app/data \
ghcr.io/your-username/mcp-explorer-x:latestOpen the App
Navigate to the Connections page.
http://localhost:8090Add a Connection
Enter your MCP server URL and transport type.
Name: My MCP Server
URL: http://localhost:3000/sse
Type: SSEExplore
Browse Tools, Prompts, Resources, or open Chat.
# Use Ctrl+K to jump anywhere
# instantly from the keyboardReady to explore your MCP servers?
One Docker command and you're live. No accounts, no setup, no friction.