MCP Server Tauri

--- title: Prompts (Slash Commands) description: Pre-built prompt templates that appear as slash commands in your AI assistant for common Tauri development tasks. head: - - meta - name: keywords content: tauri prompts, slash commands, mcp prompts, ai workflows, tauri debugging --- <script setup> import { MessageSquareCode, Bug, Zap, Settings } from 'lucide-vue-next'; </script> # <MessageSquareCode :size="28" :stroke-width="2" class="heading-icon" /> Prompts (Slash Commands) [Prompts](https://modelcontextprotocol.io/specification/2025-06-18/server/prompts) are pre-built templates that guide your AI assistant through common Tauri development tasks. In MCP clients like Claude Desktop, Cursor, and others, these appear as **slash commands** that you can invoke directly. ::: warning Prompts Not Supported Everywhere Some MCP clients don't support prompts yet (e.g., [Windsurf](https://codeium.mintlify.app/windsurf/cascade/mcp#general-information)). If slash commands don't work in your editor, see the [Getting Started guide](/guides/getting-started#manual-prompt-instructions) for a workaround. ::: ## What Are Prompts? Unlike tools (which the AI calls automatically), prompts are **user-controlled**. When you type a slash command like `/fix-webview-errors`, the prompt injects a structured set of instructions into your conversation, guiding the AI through a multi-step workflow. ::: tip When to Use Prompts Use prompts when you want a guided, multi-step workflow rather than a single action. They're perfect for debugging sessions, testing flows, and complex tasks that require multiple tools working together. ::: ## Available Prompts ### <Settings :size="20" :stroke-width="2" class="heading-icon" /> setup **Slash command:** `/setup` Guides you through setting up the MCP Bridge plugin in a Tauri project. This prompt instructs the AI to make all necessary changes to your project: 1. **Adds** the Rust crate to `src-tauri/Cargo.toml` 2. **Registers** the plugin in your app's entry point (`lib.rs` or `main.rs`) 3. **Enables** `withGlobalTauri` in `tauri.conf.json` 4. **Adds** the required permissions to your capabilities file **Example usage:** ``` /setup ``` The AI will examine your project structure and make the necessary changes. It will also verify the setup is correct and provide troubleshooting guidance if needed. ::: tip When to Use Use this prompt when you're adding the MCP bridge to a new Tauri project, or when you're not sure if your existing setup is correct. ::: --- ### <Bug :size="20" :stroke-width="2" class="heading-icon" /> fix-webview-errors **Slash command:** `/fix-webview-errors` Finds and fixes JavaScript errors in your Tauri app's webview. This prompt guides the AI through a complete debugging workflow: 1. **Connects** to your running Tauri app via the MCP bridge 2. **Retrieves** console logs and errors from the webview 3. **Analyzes** error messages, stack traces, and identifies root causes 4. **Locates** the problematic source code in your project 5. **Proposes** concrete fixes for each error found 6. **Cleans up** the session when done **Example usage:** ``` /fix-webview-errors ``` The AI will then: - Start an automation session with your app - Pull any JavaScript errors from the console - Help you understand what went wrong and how to fix it ::: warning Prerequisites - Your Tauri app must be running with the MCP bridge plugin installed - The `withGlobalTauri` option must be enabled in `tauri.conf.json` ::: **What it does behind the scenes:** | Step | Tool Used | Purpose | |------|-----------|---------| | 1 | `tauri_driver_session` | Connect to the running app | | 2 | `tauri_read_logs` | Retrieve JS errors and warnings | | 3 | Code search tools | Find source code locations | | 4 | `tauri_driver_session` | Clean up connection | ## How Prompts Work When you invoke a prompt, it sends a structured message to the AI that includes: - **Context** about what you're trying to accomplish - **Step-by-step instructions** for the AI to follow - **Tool suggestions** for each step - **Error handling guidance** for common issues The AI then executes these steps using the available MCP tools, providing feedback along the way. ## Creating Custom Workflows While the built-in prompts cover common scenarios, you can always ask your AI to perform similar workflows manually. For example: > "Connect to my Tauri app, check for any console errors, and help me fix them" This achieves the same result as `/fix-webview-errors` but gives you more control over the process. ## Prompt vs Tool: When to Use Which | Use Case | Prompt | Tool | |----------|--------|------| | Set up MCP bridge in a project | `/setup` | - | | Debug JS errors in webview | `/fix-webview-errors` | - | | Take a single screenshot | - | `tauri_webview_screenshot` | | Multi-step testing workflow | Ask AI to create one | - | | Check a specific element | - | `tauri_webview_find_element` | | Guided debugging session | Use a prompt | - | | Quick one-off action | - | Let AI choose the tool | ## See Also - [UI Automation Tools](/api/ui-automation) - Tools used by debugging prompts - [Getting Started](/guides/getting-started) - Setup instructions