Jasper AI MCP Server

The Jasper AI MCP server enables AI clients to seamlessly integrate with Jasper's powerful content generation and marketing tools. This server opens up new possibilities for developers to build AI-powered applications that leverage Jasper's advanced capabilities.

What is MCP?

MCP (Model Context Protocol) is an emerging standard that allows AI models to interact with applications through a consistent interface. It acts as an abstraction layer over HTTP, letting AI agents access application functionality without needing to understand specific API protocols.

Prerequisites

• Node.js and npm: Ensure you have Node.js (version >=20.0.0) and npm installed. You can download them from the official Node.js website.
• Jasper AI API Key: You'll need a Jasper AI API key to authenticate with the service. You can obtain one from your Jasper account settings.

Installation and Setup

1. Install the package:

   npm install @gojasper/mcp-server

2. Configure your AI client:

   Add the server configuration to your AI client's configuration file. For example, in Claude Desktop, you would modify claude_desktop_config.json:

   {
     "mcpServers": {
       "jasper": {
         "command": "npx",
         "args": ["-y", "@gojasper/mcp-server"],
         "env": {
           "JASPER_API_KEY": "<your-jasper-api-key>"
         }
       }
     }
   }

3. Restart your AI client: After saving the configuration, you may need to restart your AI client to apply the changes. You should now have access to the Jasper tools.

Available Tools

This MCP server provides a set of tools for interacting with the Jasper AI API.

| Tool Name | Description | | ------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | get-jasper-brand-voices | Retrieves a list of defined brand voices/tones from your Jasper account. | | get-jasper-style-guides | Retrieves every style guide in your Jasper workspace, including each guide's name, ID, description, custom instructions, custom terms, and enabled grammar rules. | | get-jasper-audiences | Retrieves a list of defined audiences from your Jasper account. | | get-jasper-products | Retrieves products from your Jasper workspace with search and pagination, each with its full set of attributes. | | search-knowledge-base | Searches your Jasper knowledge base for relevant documents and information. | | get-jasper-agents | Retrieves a list of available jasper agents with pagination and filtering options. | | run-jasper-agent | Executes a jasper agent to generate structured content using predefined templates. | | generate-content | Generates content based on a given command, with options for specifying one or multiple audiences, tone, style, products, and background knowledge. Passing a styleId applies that style guide's rules during generation. | | upload-attachment | Uploads a local file, raw text, or a URL as a temporary attachment and returns a tmp_ ID to pass as attachmentIds on a generation. Nothing is added to your Knowledge Base. Local file paths require the locally-installed server. | | get-jasper-geo-analyses | Lists the GEO analyses in the workspace — each tracked website with its latest completed run. | | get-jasper-geo-run | Retrieves selected sections of a GEO run, or the complete run when sections is omitted. |

The GEO tools are read-only and are registered only when GEO_REPORTS_ENABLED=true.

Example Usage

Here are some example prompts you can use with an AI client that has the Jasper MCP server configured:

• "Generate a blog post about the benefits of using AI in marketing."
• "Rewrite the following paragraph using our 'Marketing' style guide: [your paragraph here]"
• "Rewrite the following paragraph for [target audience]: [your paragraph here]"
• "What are the key points from our last marketing meeting? (searches knowledge base)"
• "List all available audiences for our marketing campaigns."
• "Write a launch email using the messaging in ~/Desktop/positioning.pdf" (uploads the file as one-time context)