Access Your AI Prompts Directly from Claude: Sehla's MCP Integration

Why MCP Integration Matters for Your Workflow

If you're using AI assistants like Claude for development, you've probably experienced this frustration: you're deep in a coding session, need a specific prompt template, and have to leave your IDE to search for it in a browser tab.

That context switch kills momentum.

Sehla's Model Context Protocol (MCP) integration solves this by bringing your entire prompt library directly into your development environment. Search, access, and create prompts without ever leaving Claude Code, VSCode, or other MCP-enabled tools.

What Is Model Context Protocol?

The Model Context Protocol is an open standard that enables AI assistants to connect with external tools and data sources. Think of it as a bridge between your AI assistant and the services you use—in this case, your Sehla prompt library.

With MCP, Claude can:

  • Search your private prompt collection
  • Retrieve specific prompts on demand
  • Create new prompts from successful conversations
  • Access your folder structure and organization
  • Handle dynamic variables automatically

How Sehla's MCP Integration Works

Sehla provides a full-featured MCP server that exposes six core tools to AI assistants:

1. Search Prompts

Find exactly what you need with full-text search across titles, descriptions, and content. Claude can search your entire library based on natural language queries.

2. Get Prompt Details

Retrieve complete prompt content with metadata, tags, and automatic variable extraction. Perfect for grabbing templates mid-conversation.

3. Create New Prompts

Save successful AI interactions as reusable prompts directly from your IDE. Sehla automatically detects dynamic variables (like <<topic>> or <<audience>>) and organizes content properly.

4. List Folders

Browse your hierarchical folder structure to understand how prompts are organized. See ownership, sharing status, and permissions at a glance.

5. Create Folders

Organize new prompts into folders without switching tools. Maintain your organizational system as you work.

6. Get User Info

Check your current subscription limits, usage statistics, and account details—helpful for understanding what you can create.

Setting Up MCP with Claude

Getting started takes less than a minute:

For Claude Code:

claude mcp add --transport http sehla https://api.sehla.ai/mcp

For VSCode: Create a .vscode/mcp.json file:

{
"servers": {
"sehla": {
"url": "https://api.sehla.ai/mcp",
"type": "http"
}
}
}

Once configured, authenticate with your existing Sehla account via Google OAuth. Your AI assistant gains secure access to your prompt library with the same permissions you have in the web app.

Real-World Use Cases

Scenario 1: The Template Hunter You're building API documentation and need your standard API endpoint template. Instead of opening Sehla in a browser, you ask Claude: "Search my prompts for API documentation template." Claude retrieves it instantly with all variables ready for customization.

Scenario 2: The Saver You've just had a productive conversation with Claude about writing marketing emails. The prompt structure worked perfectly. Instead of manually copying it, you tell Claude: "Save this conversation as a new prompt called 'Email Marketing Framework' in my Marketing folder." Done.

Scenario 3: The Team Player Your team lead shared a folder of approved prompts for customer support responses. You access them directly through Claude without leaving your chat interface—maintaining consistency while staying in flow.

Technical Implementation Highlights

Sehla's MCP server is built on MCP-Nest, providing:

  • OAuth 2.1 authentication with Google
  • HTTP + SSE transport for broad compatibility
  • JWT token security (7-day expiry, same as web app)
  • Rate limiting and error handling
  • Automatic variable detection using <<variableName>> syntax
  • Shared content access respecting your permissions

The server runs at https://api.sehla.ai/mcp and follows RFC 8414 and RFC 9728 standards for OAuth discovery and protected resource metadata.

Benefits You'll Actually Notice

No Context Switching

Stay in your IDE or AI assistant. Your prompts come to you.

Consistent Organization

Folders, tags, and sharing all work the same way they do in the Sehla web app. No learning curve.

Dynamic Variables Built-In

Prompts with variables like <<customerName>> or <<productType>> are automatically parsed. Claude prompts you for values before using the template.

Version History Included

Every prompt you access includes its version history. Restore previous versions if needed (based on your subscription plan).

Secure by Default

Your authentication tokens follow the same security model as the web app. Only you (and people you've explicitly shared with) can access your prompts.

Subscription Considerations

MCP integration works with both Free and Pro plans:

Free Plan:

  • Access up to 50 prompts
  • 5 version history entries per prompt
  • Full MCP search and retrieval

Pro Plan ($4.99/month):

  • Unlimited prompts and folders
  • Unlimited version history
  • Folder sharing with teams
  • Enhanced organization capabilities

All MCP features respect your plan limits. If you're approaching limits, Claude will inform you—or you can check usage with the get-user-info tool.

Getting Started Today

Ready to stop context switching?

  1. Have a Sehla account? Set up MCP in Claude Code or VSCode using the commands above
  2. New to Sehla? Sign up for free to create your first prompt library
  3. Questions? Check our MCP documentation for advanced configuration

The Model Context Protocol is transforming how developers interact with AI tools. With Sehla's MCP integration, your prompt library becomes a first-class citizen in your development workflow—accessible, searchable, and always in sync.

Start building your MCP-connected prompt library today at sehla.ai.