Vaali

Vaali MCP Server enables interactive AI agent workflows by combining official MCP elicitation with intelligent parameter guidance. It supports both local, stdio-based operation for development and HTTP/SSE-based deployment for cloud use, enabling robust parameter collection, tool usage, and workflow prompts across clients.

How to use

You connect a client to Vaali to access tools, prompts, and elicitation capabilities. You can run the server locally in stdio mode for development or use an HTTP/SSE transport to connect from cloud-based MCP clients. The server automatically adapts to the client’s capabilities, providing interactive forms when possible and offering robust guidance when forms are not available. Use natural language prompts to ask for weather, perform calculations, or generate reports, while the server coordinates parameter collection and tool calls to deliver the final result.

How to install

Follow these steps to set up Vaali MCP Server on your machine and prepare for development or testing.

# Prerequisites
# Node.js 18+ is required
# Optionally install VS Code for development

# 1. Clone the project
# Replace <repository-url> with the actual repository URL
git clone <repository-url>
cd vaali

# 2. Install dependencies
npm install

# 3. Build the project
npm run build

# 4. Run tests (optional but recommended)
npm run test:advanced-concept
npm run test:working-advanced
npm run test:all

# 5. Start in STDIO mode for local Claude Desktop testing
npm run start:stdio

# 6. Start in SSE mode for web-based transport testing
npm run start:sse

Configuration and startup for local development

To integrate with Claude Desktop during development, configure Claude Desktop to run Vaali as a local MCP server via STDIO. Provide an absolute path to the server script and set the working directory accordingly.

{
  "mcpServers": {
    "vaali": {
      "command": "node",
      "args": ["C:/absolute/path/to/vaali/lib/src/index.js", "stdio"],
      "cwd": "C:/absolute/path/to/vaali",
      "env": {
        "NODE_ENV": "development"
      }
    }
  }
}

Usage with HTTP/SSE transport

When you deploy Vaali to the cloud or use an HTTP/SSE transport, connect MCP clients to the SSE endpoint to receive messages and send prompts. The SSE endpoint is exposed after deployment and is used by web-based clients and SDK transports.

// Example: Connecting to an HTTP/SSE endpoint from an MCP client
// This snippet shows the transport initialization using a placeholder URL
const { SSEClientTransport } = require('@modelcontextprotocol/sdk/client/sse.js');
const { Client } = require('@modelcontextprotocol/sdk/client/index.js');

const client = new Client(
  { name: "vaali-client", version: "1.0.0" },
  { capabilities: {} }
);

const transport = new SSEClientTransport(
  new URL('https://your-app-name.azurewebsites.net/sse')
);

await client.connect(transport);

What to expect from the server

Vaali provides resources, tools, prompts, and elicitation capabilities. It supports both official MCP elicitation for structured data collection and universal parameter guidance for broader compatibility. The system adapts to client capabilities, offering interactive forms when available and seamless fallbacks when necessary.

Azure deployment and hosting notes

You can deploy Vaali to Azure App Service to gain cloud access and scalability. The recommended deployment method uses GitHub Actions to build, test, and deploy automatically.

Important: Make sure to configure the deployment environment to enable TypeScript compilation, install production dependencies, and set the proper Node.js runtime. The production URL and SSE endpoint are exposed after deployment.

Security and performance considerations

Use HTTPS for all endpoints, enable CORS for web access, and avoid logging sensitive data. Choose an appropriate App Service Plan (B1 for demos or S1 Standard for production) and enable auto-scaling if needed. Maintain a health check endpoint for reliable monitoring.

Testing and troubleshooting

Run the full test suite to validate advanced elicitation concepts and working advanced scenarios. If you encounter transport issues, verify the transport mode (stdio vs SSE) and ensure the endpoint URLs are reachable from your client environment.

Available tools

Weather Tools

Current conditions, forecasts, and location lookup to support weather-related tasks.

Analysis Tools

Text analysis, calculations, and data processing utilities.

Elicitation Tools

Interactive parameter collection testing and validation.