# Mcp Chat Studio > A powerful MCP testing tool with multi-provider LLM support (Ollama, OpenAI, Claude, Gemini). Test, debug, and develop MCP servers with a modern UI. - **Type:** MCP server - **Install:** `agentstack add mcp-joecastrom-mcp-chat-studio` - **Verified:** Yes β€” security-reviewed for prompt injection and unsafe behavior - **Seller:** [JoeCastrom](https://agentstack.voostack.com/s/joecastrom) - **Installs:** 0 - **Category:** [AI & ML](https://agentstack.voostack.com/c/ai-and-ml) - **Latest version:** 0.1.0 - **License:** MIT - **Upstream author:** [JoeCastrom](https://github.com/JoeCastrom) - **Source:** https://github.com/JoeCastrom/mcp-chat-studio ## Install ```sh agentstack add mcp-joecastrom-mcp-chat-studio ``` Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent. ## About # πŸš€ MCP Chat Studio **The Ultimate Testing Platform for MCP Servers** Test, debug, and develop Model Context Protocol servers with a beautiful glassmorphism UI and a powerful Workspace mode. **Record scenarios** β€’ **Collections + run reports** β€’ **Workspace canvas** β€’ **Workflow export (Python + Node)** β€’ **Mock servers** β€’ **Docs generator** β€’ **Zero config with Ollama** **Need help?** Use the **Studio Assistant** (bottom‑right compass) to learn features, generate workflows, import OpenAPI, and navigate the UI β€” just ask it. [](https://github.com/JoeCastrom/mcp-chat-studio/actions/workflows/ci.yml) [](https://mseep.ai/app/joecastrom-mcp-chat-studio) > **πŸ†• Major Update (January 2025):** This version includes significant new features like OAuth config UI, LLM settings persistence, E2E tests, session management, security enhancements, and much more. The previous stable version is available in the [`v2` branch](https://github.com/JoeCastrom/mcp-chat-studio/tree/v2) as a backup. ## 🎬 Showcase | Workspace Canvas | Studio Assistant | AI Workflow Builder | Documentation Generator | | ------------------------------------------------------- | ------------------------------------------------------ | ----------------------------------------------------------------- | ---------------------------------------------------------------- | | | | | | > **πŸ“Œ Local MCP Test Bench** - Designed for development. Not intended for internet exposure. --- ## ⚑ Get Started in 60 Seconds ```bash # 1. Clone and install git clone https://github.com/JoeCastrom/mcp-chat-studio.git && cd mcp-chat-studio && npm install # 2. Start (works with Ollama out of the box - no API keys needed!) npm run dev # 3. Open http://localhost:3082 and start testing! πŸŽ‰ ``` **That's it!** You now have a complete MCP testing platform running locally. Toggle **Classic / Workspace** in the header to pick your workflow. Use the **model badge** in the header to switch LLM providers; the **βš™οΈ Settings** button handles model/auth tuning. --- ## πŸ†š Why Not Just Use Claude Desktop or ChatGPT? | Feature | MCP Chat Studio | Claude Desktop | ChatGPT | | -------------------------------- | :-------------: | :--------------: | :--------------: | | Test ALL tools automatically | βœ… | ❌ | ❌ | | Record & Replay scenarios | βœ… | ❌ | ❌ | | Workflow debugger (breakpoints) | βœ… | ❌ | ❌ | | Bulk testing (multiple inputs) | βœ… | ❌ | ❌ | | Response diffing | βœ… | ❌ | ❌ | | Contract testing | βœ… | ❌ | ❌ | | Tool usage analytics | βœ… | ❌ | ❌ | | Custom assertions (14 operators) | βœ… | ❌ | ❌ | | Low-level MCP debugging | βœ… | ❌ | ❌ | | Mock server generator | βœ… | ❌ | ❌ | | 9 LLM providers + custom | βœ… | ❌ (Claude only) | ❌ (OpenAI only) | | Local/no API keys needed | βœ… (Ollama) | ❌ | ❌ | | Multi-environment profiles | βœ… | ❌ | ❌ | | Session branching | βœ… | ❌ | ❌ | --- ## 🧭 Two Layouts: Classic + Workspace **Classic Mode** keeps the familiar sidebar + tabs for focused work. **Workspace Mode** turns every tool into a draggable panel on an infinite canvas: - **Radial menu** to add panels (right-click) - **Quick access bar** for fast panel focus - **Zoom, pan, mini-map, fit-all** - **Command palette** (Ctrl+K / Ctrl+Shift+P) - **Workspace sessions** + **export/import** bundles --- ## 🎨 Highlights: Advanced Testing & Debugging Suite ### πŸ› Workflow Debugger Debug complex workflows with professional debugging tools: - **Breakpoint debugging** - Pause execution at any node - **Step-through mode** - Execute one node at a time - **Variable inspection** - View inputs, outputs, and context - **Session management** - Pause, resume, step, or abort ### ✨ AI Workflow Builder Generate workflows from a natural-language goal: - **Tool-aware prompts** - Uses connected MCP tools as context - **One-click generate** - Creates a workflow you can edit and run - **Works with Workspace and Classic modes** - **Preflight validation** - Catch missing tool args and invalid JSON before execution ### πŸ“¦ Workflow Export Turn visual flows into runnable scripts: - **Python (mcp SDK)** and **Node.js (mcp SDK)** - **Tool calls + variable substitution** baked in - **Shareable** automation for CI or docs ### πŸ” Advanced Inspector Three powerful new testing tools: - **πŸ“Š Timeline** - Chronological log of all JSON-RPC messages - **πŸ§ͺ Bulk Test** - Execute tools with multiple inputs (parallel/sequential) - **πŸ”₯ Bulk Test Heatmap** - Visualize latency spread across inputs + export JSON with heatmap data - **🎬 Fuzz Fail β†’ Scenario** - Save failing cases as replayable scenarios - **πŸ“š Failure Datasets** - Store bulk-test failures as dataset rows for data runs - **πŸ”€ Diff** - Side-by-side comparison with similarity scoring - **🌐 Cross-Server Snapshot** - Run one tool across servers and diff output - **🧭 Matrix wizard** - Pick servers + common-tool filter before running diffs - **🌐 One-click Matrix** - Run a tool across servers from the Inspector response - **πŸ•˜ History β†’ Matrix** - Launch a cross-server compare directly from past runs - **πŸ§ͺ Schema Fuzzing** - Generate edge-case inputs from tool schemas - **πŸ” Resolved Preview** - See variable substitutions before execution - **πŸ” Per-request Auth Overrides** - Bearer/Basic/API key + header/query overrides for OpenAPI proxy tools - **πŸ” OAuth Settings UI** - Configure OAuth providers without touching config.yaml or .env (includes a Test OAuth button) ### πŸ“‹ Contract Testing Consumer-driven contract testing for MCP servers: - **Define contracts** - Specify expected tool behavior - **Multiple assertions** - Schema, contains, equals, response time, custom - **Auto-generation** - Generate contracts from tool schemas - **Version tracking** - Track contract changes over time - **Schema Watch** - Live drift detection with optional background checks - **Schema CI Gate** - One-click baseline export + CI fail-on-change command ### πŸ“Š Tool Explorer & Analytics Real-time usage statistics and performance metrics: - **Usage tracking** - Calls, success rates, latency per tool - **Performance metrics** - p50/p95/p99 latency tracking - **Error monitoring** - Recent errors per tool - **Server health badges** - Connected/Failed/Not connected with details + retry - **Leaderboards** - Most-used tools across servers - **Health dashboard** - System-wide health overview - **Flake radar** - Highlights tools with rising failure or latency jitter - **Flake alerts** - Baseline + regression checks for tool reliability ### πŸ“š Documentation Generator Publish MCP server docs in seconds: - **Markdown / HTML / JSON** - **Tools + schemas + examples** - **Shareable docs bundle** ### ⏱️ Monitors & Performance Keep an eye on quality and latency: - **Scheduled monitors** with pass/fail history - **Performance tab** with p50/p95/p99 and trends - **Health check dashboards** ### 🧭 Workspace Mode Build your own testing cockpit with floating panels: - **Add panels fast** - Radial menu + quick-access bar - **Canvas controls** - Zoom, pan, mini-map, fit-all - **Command palette** - Ctrl+K / Ctrl+Shift+P - **Sessions + bundles** - Save/restore layouts, export/import JSON - **Templates** - Save and reuse workspace presets --- ## 🎨 Visual MCP Server Generator Create production-ready MCP servers without writing boilerplate code! - 🎨 **Visual tool designer** with parameters and types - 🧾 **OpenAPI import** (JSON + YAML) to auto-create tools from API specs (paths + webhooks) - πŸ›°οΈ **OpenAPI Proxy mode** - generate runnable MCP servers that call real APIs - πŸ“¦ **Project bundle export** - download a ready-to-run MCP server folder from the Generator - πŸ—œοΈ **ZIP export** - one-click runnable project archive - πŸ§ͺ **Test in Studio** - copy config + open Add Server to wire the generated MCP quickly - βœ… **Guided Test flow** - step-by-step run commands + required working directory - πŸš€ **Run & Connect (Auto)** - Studio writes a temporary project folder and connects it for you (Python auto-run requires `mcp` installed) - Prompts to import selected OpenAPI endpoints if no tools are in the designer yet - Auto-suffixes the server name to avoid clobbering existing configs (e.g. `my-mcp-server-auto-7f3a`) - πŸ“‚ **Save to Folder** - write the project files directly to a local folder (supported browsers) - πŸ” **Auth mapping** - OpenAPI security schemes flow into MCP tool metadata - 🌐 **HTTP hints** - Imported endpoints carry read-only/destructive hints via tool annotations - 🧭 **Server URLs + tags** - Choose base URL from spec servers[] and bulk-select endpoints by tag - ⭐ **Sample OpenAPI loader** - one-click Petstore to explore the workflow - 🐍 **Python (mcp SDK)** and πŸ“¦ **Node.js (TypeScript)** code generation - ⚑ Copy to clipboard or download instantly - πŸ”§ Perfect for prototyping and teaching MCP --- ## 🎯 Perfect For - **πŸ”§ MCP Server Developers** - Test your tools without manual clicking - **πŸ€– AI App Builders** - Compare GPT-4 vs Claude vs Llama on the same tasks - **🏒 Enterprise Teams** - Share test scenarios and environments - **πŸ› Debugging** - See raw MCP protocol messages when things break - **πŸ“š Learning MCP** - Visual tool designer teaches protocol structure --- ## ✨ Why MCP Chat Studio? - **🎯 Built for MCP Development** - Test and debug MCP servers without writing code - **πŸ”§ 9 LLM Providers + Custom** - Ollama, OpenAI, Claude, Gemini, Azure, Groq, Together AI, OpenRouter + Custom - **🧠 Ollama model picker** - UI dropdown of locally installed models (auto-detected) - **πŸ”„ Provider Switcher** - Swap LLMs from the model badge and manage visible providers in one place - **🧭 Studio Assistant** - Floating help chat aware of your current panel/layout, with dock + pop‑out modes, quick actions, FAQ fallback, and drag‑and‑drop OpenAPI import - Commands list + recent command chips + safe confirmations + workspace builder (add/close/resize/arrange panels, sessions, export/import) - Paste OpenAPI URLs/JSON or upload specs to auto-import into the Generator - Auto action button to **Generate + Test** immediately after import - Resizable panel with dock, pop‑out, and size toggle - **πŸ’¬ Smart chat import** - Pasting an OpenAPI URL/JSON in the main chat offers to import into Generator - **πŸ§ͺ Test Scenarios** - Record, replay, and validate tool executions - **🎬 History β†’ Scenario** - Turn real tool calls into replayable test flows - **πŸ” Re-run + Diff** - Execute any past tool call and compare outputs instantly - **🌐 Matrix Runs** - Execute the same scenario across multiple servers with baseline diffs - **πŸ“Š Data Runs** - Replay scenarios against a JSON dataset - **πŸ“š Dataset Library** - Save reusable data tables for scenarios - **πŸ” OAuth in UI** - Configure auth providers and session-based servers from the app - **🎬 Fuzz Fail β†’ Scenario** - Convert bulk test failures into replayable scenarios - **🌐 Cross-Server Snapshots** - Compare live tool outputs across servers - **⚑ Flake Radar** - Spot flaky tools using live failure + jitter signals - **🚨 Flake Alerts** - Detect regressions against a saved reliability baseline - **πŸ“š Collections & Run Reports** - Batch scenarios, run iterations, export JSON/JUnit + CI gate - **⭐ Golden Baselines** - Mark trusted runs and diff against them - **πŸ“Έ Run Snapshots** - Deterministic replay + drift checks on collections (export/import) - **🚦 Drift Gate** - Export pass/fail gate summary for CI - **🧭 Workspace Mode** - Floating panels, zoom, mini-map, and command palette - **πŸ“Š Response Diffing** - Semantic JSON comparison with color-coded changes - **πŸ“‹ Schema Validation** - Contract testing with auto-inferred schemas - **πŸ”” Schema Watch** - Detect live schema drift across connected servers - **πŸ“¦ Workflow Export** - Generate Python or Node.js scripts from workflows - **πŸ“š Docs Generator** - Publish MCP documentation in one click - **πŸ“¦ Project Bundles** - Export/import full test setups (collections, mocks, envs) - **πŸ” Custom Assertions** - 14 operators with JSONPath support - **πŸ” Low-Level Debugging** - Inspector tab for raw MCP protocol inspection - **🐳 Production Ready** - Docker support, CI/CD, security-hardened - **πŸ’‘ Zero Config Start** - Works with Ollama out of the box, no API keys needed - **🎨 Beautiful UI** - Modern glassmorphism design with dark/light themes --- ## πŸ“Έ Screenshots Click to view more screenshots ### Light Mode ### Adding MCP Servers ### Tool Inspector (Low-Level Debugging) ### Tool Testing ### Keyboard Shortcuts --- ## πŸš€ Features ### πŸ’¬ Multi-Provider LLM Chat - **Ollama** - Local LLM (llama3, mistral, qwen, etc.) - **OpenAI** - GPT-4o, GPT-4, GPT-3.5 - **Anthropic** - Claude 3.5, Claude 3 - **Google Gemini** - Gemini Pro, Gemini Flash - **Azure OpenAI** - Enterprise Azure deployments - **Groq** - Ultra-fast inference (Mixtral, LLaMA) - **Together AI** - Open-source models - **Custom LLM** - OpenAI-compatible endpoints with optional OAuth token fetch - **Real-time streaming** with typing effect ### πŸ”§ MCP Tool Management - **Dynamic server management** - Add/remove servers at runtime - **STDIO & SSE transports** supported - **Environment variables** - Configure API keys, URLs per server - **Import YAML/JSON** - Paste config from docs - **Config preview** - See generated config before adding ### πŸ§ͺ Tool Testing - **Test All Tools** - Smoke test all connected tools - **Response preview** - See what each tool returns - **Timing** - Measure tool response times - **Safe mode** - Skip risky tools (Click, Type, Launch) - **Error detection** - Uses MCP's `isError` field ### πŸ”§ Inspector Tab (Low-Level Debugging) - **Manual tool execution** - Call any tool with custom arguments - **Auto-generated forms** - Input fields from JSON schema - **Protocol log** - See raw MCP request/response JSON - **SSE event viewer** - Real-time server events for SSE transports - **Resources/Prompts API** - Full MCP protocol support - **Timeline** - Chronological JSON-RPC message log with filtering - **Bulk Testing** - Execute tools with arrays of inputs - **Diff Tool** - Side-by-side result comparison with similarity % ### πŸ› Workflow Debugger - **Breakpoint debugging** - Set breakpoints on any workflow node - **Step-through execution** - Execute workflows one node at a time - **Variable inspection** - View all context, inputs, and outputs - **Pause/Resume/Abort** - Full execution control - **Debug sessions** - Multiple concurrent debug sessions - **Execution log** - Complete history of node executions - **11 API endpoints** - Full REST API for debugging ### πŸ“‹ Contract Testing Suite - **Define contracts** - JSON-based contract definitions - **Schema assertions** - Validate response structure - **Response time limits** - Performance SLAs - **Custom assertions** - Path-based queries with operators - **Auto-generation** - Create contracts from tool schemas - **Schema Watch** - Live drift detection with on-demand or auto checks - **Version tracking** - Contract versioning and updates - **CRUD API** - Full contract lifecycle management - **Test reporting** - Detailed pass/fail with error messages ### πŸ“Š Tool Explorer & Analytics - **Usage statistics** - … ## Source & license This open-source MCP server is cataloged on AgentStack and links to its original source β€” we do not rehost the code. - **Author:** [JoeCastrom](https://github.com/JoeCastrom) - **Source:** [JoeCastrom/mcp-chat-studio](https://github.com/JoeCastrom/mcp-chat-studio) - **License:** MIT Install and usage instructions live in the source repository linked above. ## Pricing - **Free** β€” Free ## Security capabilities Automated source analysis of v0.1.0 β€” what this tool can access: - **Network access:** yes - **Filesystem access:** no - **Shell / process execution:** yes - **Environment & secrets:** yes - **Dynamic code execution:** no *"Yes" means the capability is present in the source β€” more access means more to trust, not that it is unsafe.* ## Versions - **0.1.0** β€” security scan: passed β€” Imported from the upstream source. ## Links - Listing page: https://agentstack.voostack.com/l/mcp-joecastrom-mcp-chat-studio - Seller: https://agentstack.voostack.com/s/joecastrom - Browse the marketplace: https://agentstack.voostack.com/browse --- Listed on AgentStack β€” the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.