MCP Bundles

https://www.mcpbundles.com/docs
Admin
MCP Bundles connects AI assistants like ChatGPT, Claude, and Cursor to real-world tools and APIs,...
Tokens:90,864
Snippets:1,000
Trust Score:5.4
Update:7 months ago
Show doc for...
Tokens:
Raw
### Example Prompts for Using MCP Tools in Cursor

Source: https://www.mcpbundles.com/docs/integrations/cursor

Examples of prompts that leverage MCP tools within Cursor for various development tasks, such as fetching data, automating deployment, and CRM integration.

```text
"Get the latest campaign data from Smartlead and create a TypeScript interface for it"
```

```text
"List my GitHub repos, find the one called 'api-server', and check if there are any open PRs"
```

```text
"Fetch the latest 10 contacts from HubSpot and generate a SQL migration to import them"
```

--------------------------------

### Manual MCP Configuration File Example

Source: https://www.mcpbundles.com/docs/integrations/vscode

This JSON snippet shows how to manually configure MCP servers in the global or workspace configuration files. It includes examples for adding multiple bundles with different URLs.

```json
{
  "mcpServers": {
    "marketing": {
      "url": "https://mcp.mcpbundles.com/bundle/abc123"
    },
    "dev-tools": {
      "url": "https://mcp.mcpbundles.com/bundle/def456"
    }
  }
}
```

--------------------------------

### Install mcpbundles-proxy on Windows

Source: https://www.mcpbundles.com/docs/how-to/desktop-connection

Installs the mcpbundles-proxy using pip on Windows. It also provides instructions on how to set it up to run automatically at startup.

```bash
pip install mcpbundles-proxy

# Run at startup:
# 1. Press Win + R
# 2. Type shell:startup
# 3. Create shortcut to: mcpbundles-proxy start

# Logs location:
# %USERPROFILE%\.mcpbundles\logs\proxy.log
```

--------------------------------

### Minimal MCP Configuration: Single Bundle

Source: https://www.mcpbundles.com/docs/integrations/cursor

A basic JSON configuration for connecting a single MCP bundle to Cursor. This example shows the minimal setup required to specify the bundle's URL.

```json
{
  "mcpServers": {
    "marketing": {
      "url": "https://mcp.mcpbundles.com/bundle/abc123"
    }
  }
}
```

--------------------------------

### Example Prompts for Gemini CLI

Source: https://www.mcpbundles.com/docs/integrations/gemini-cli

Demonstrates how to phrase natural language prompts for the Gemini CLI to interact with different services. These examples cover data retrieval, GitHub operations, and CRM tasks.

```text
"Check my Smartlead campaigns and show the top 3 by reply rate"
"List my GitHub repos and tell me which ones have the most stars"
"Add John Doe (john@example.com) to HubSpot as a new contact"
```

--------------------------------

### Install mcpbundles-proxy on Linux

Source: https://www.mcpbundles.com/docs/how-to/desktop-connection

Provides installation commands for mcpbundles-proxy on Debian/Ubuntu and Fedora/RHEL-based Linux distributions using pipx.

```bash
# Ubuntu/Debian
sudo apt install pipx
pipx install mcpbundles-proxy

# Fedora/RHEL
sudo dnf install pipx
pipx install mcpbundles-proxy
```

--------------------------------

### Example Prompts for AI Tool Usage

Source: https://www.mcpbundles.com/docs/how-to/connecting-your-ai

These are example natural language prompts that can be used to instruct AI models like ChatGPT or Claude to utilize integrated tools for various tasks. The AI will interpret these requests and automatically call the relevant functions or APIs.

```text
"Check my Smartlead campaigns and tell me which has the best reply rate"

"Add these people to HubSpot: John Doe (john@example.com), Jane Smith (jane@example.com)"

"List my GitHub repos and check for open PRs"
```

```text
"Can you check my email campaigns and summarize performance?"

"Create a contact in my CRM for Tony Lewis at tony@example.com"

"What GitHub repositories do I have access to?"
```

```text
"Get the latest campaign data from Smartlead and create a TypeScript interface for it"

"List my GitHub repos, find 'api-server', and check for open PRs"

"Fetch the latest 10 contacts from HubSpot and generate SQL for importing them"
```

--------------------------------

### Configure mcpbundles-proxy to Run at Startup on macOS

Source: https://www.mcpbundles.com/docs/how-to/desktop-connection

Creates and loads a LaunchAgent plist file to automatically start the mcpbundles-proxy when the user logs in on macOS.

```bash
# Create LaunchAgent
cat > ~/Library/LaunchAgents/com.mcpbundles.proxy.plist << 'EOF'
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
<plist version="1.0">
<dict>
    <key>Label</key>
    <string>com.mcpbundles.proxy</string>
    <key>ProgramArguments</key>
    <array>
        <string>/usr/local/bin/mcpbundles-proxy</string>
        <string>start</string>
    </array>
    <key>RunAtLoad</key>
    <true/>
</dict>
</plist>
EOF

# Load the service
launchctl load ~/Library/LaunchAgents/com.mcpbundles.proxy.plist
```

--------------------------------

### AI Assistant - Getting Help with Inputs

Source: https://www.mcpbundles.com/docs/guides/user/bundle-studio

Illustrates how the AI assistant can guide users in finding necessary inputs for a tool, such as suggesting a prerequisite tool to retrieve required IDs.

```text
You: I want to test get_campaign_stats but I don't know my campaign ID

AI: You can use the list_campaigns tool first to see all your campaigns
and their IDs. Would you like me to walk you through that?
```

--------------------------------

### Creating MCPBundles with Anthropic CLI

Source: https://www.mcpbundles.com/docs/concepts/mcpb-files

These commands demonstrate how to install the Anthropic CLI tool, navigate to your server directory, initialize a manifest file, and package your project into an .mcpb file.

```bash
# Install the tool
npm install -g @anthropic-ai/mcpb

# Go to your MCP server directory
cd your-mcp-server/

# Create the manifest
mcpb init

# Package everything
mcpb pack
```

--------------------------------

### MCP Configuration: With Other MCP Servers

Source: https://www.mcpbundles.com/docs/integrations/cursor

An example of an MCP configuration that integrates MCPBundles with other MCP servers, such as local-filesystem and github, enabling broader access to system resources.

```json
{
  "mcpServers": {
    "mcpbundles-marketing": {
      "url": "https://mcp.mcpbundles.com/bundle/abc123"
    },
    "local-filesystem": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-filesystem", "/Users/you/projects"]
    },
    "github": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-github"]
    }
  }
}
```

--------------------------------

### Troubleshooting: JSON Syntax Errors in mcp.json

Source: https://www.mcpbundles.com/docs/integrations/cursor

Illustrates common JSON syntax errors, specifically missing and trailing commas, within the mcp.json configuration file. Provides corrected examples for proper formatting.

```json
{
  "mcpServers": {
    "bundle1": { "url": "..." }  // ❌ Missing comma
    "bundle2": { "url": "..." }
  }
}
```

```json
{
  "mcpServers": {
    "bundle1": { "url": "..." },  // ✅ Has comma
    "bundle2": { "url": "..." }
  }
}
```

```json
{
  "mcpServers": {
    "bundle": { "url": "..." },  // ❌ Trailing comma
  }
}
```

```json
{
  "mcpServers": {
    "bundle": { "url": "..." }  // ✅ No trailing comma
  }
}
```

--------------------------------

### Test MCP Bundle Tools in AI

Source: https://www.mcpbundles.com/docs/getting-started

These examples show how to interact with your integrated MCP bundle tools from within your AI chat interface. First, you ask the AI to list available tools, and then you execute a specific tool, such as `list_campaigns`, to retrieve data.

```text
What MCP tools do you have available?
Use the list_campaigns tool to show my email campaigns
```

--------------------------------

### Add MCP Bundle using Claude Code CLI

Source: https://www.mcpbundles.com/docs/getting-started

This command-line interface (CLI) example demonstrates how to add an MCP bundle to Claude using the `claude` command. It first adds the bundle with a specified name and URL, then starts a session and authenticates the tools.

```shell
claude mcp add --transport http my-bundle https://mcp.mcpbundles.com/bundle/your-bundle-slug
claude  # Start session
/mcp    # Authenticate tools
```

--------------------------------

### Install mcpbundles-proxy on macOS with Homebrew

Source: https://www.mcpbundles.com/docs/how-to/desktop-connection

Installs the mcpbundles-proxy using Homebrew on macOS. It first installs pipx, a tool for installing and running Python applications in isolated environments.

```bash
brew install pipx
pipx install mcpbundles-proxy
```

--------------------------------

### Configuring Environment-Specific Bundles

Source: https://www.mcpbundles.com/docs/integrations/gemini-cli

Demonstrates how to configure environment-specific bundles within the `~/.gemini/settings.json` file. This allows for different bundles to be used for different environments like production or staging.

```json
{
  "mcpServers": {
    "prod": {
      "command": "npx",
      "args": ["mcp-remote", "https://mcp.mcpbundles.com/bundle/prod-id"]
    }
  }
}
```

--------------------------------

### AI Assistant - Understanding a Tool

Source: https://www.mcpbundles.com/docs/guides/user/bundle-studio

Example conversation demonstrating how to ask the AI assistant for information about a specific tool, including its purpose and required/optional inputs.

```text
You: What does the create_lead tool do?

AI: The create_lead tool adds a new lead to your Smartlead campaign.
It needs:
- email (required)
- first_name (optional)
- last_name (optional)
- campaign_id (required)

Would you like me to help you test it?
```

--------------------------------

### Automated Code Review and Risk Assessment Prompt

Source: https://www.mcpbundles.com/docs/use-cases/developer-productivity

This prompt guides an AI to perform a comprehensive code review for a pull request. It focuses on summarizing changes, identifying potential issues (security, performance, maintainability), verifying test coverage, flagging breaking changes, and suggesting improvements with code examples. It aims to provide consistent and thorough feedback, freeing up senior engineers.

```text
"Review this pull request: [PR URL or number]
- Summarize all code changes and their purpose
- Identify potential security, performance, or maintainability issues
- Check if tests adequately cover the changes
- Flag any breaking changes or migration requirements
- Suggest specific improvements with code examples"
```
Tokens:
Raw