Skip to main content
Glama
joaomj

deep-search-mcp

by joaomj
README.md
# Deep Search MCP Server

A deep web search MCP server using LinkUp API.

This is a TypeScript-based MCP server that implements deep web search capabilities. It demonstrates core MCP concepts by providing:

- Tools for performing deep web searches
- Structured results from LinkUp API

## Features

### Tools
- `deep_search` - Perform deep web searches
  - Takes query string as required parameter
  - Optional max_results parameter (default: 5)
  - Returns structured search results

## Setup

1. Install dependencies:
```bash
npm install
```

2. Build the server:
```bash
npm run build
```

3. Configure the MCP server in your settings:
```json
{
  "mcpServers": {
    "deep-search-mcp": {
      "command": "node",
      "args": ["/home/joao/Cline/MCP/linkup-mcp-server/build/index.js"],
      "env": {
        "LINKUP_API_KEY": "your-api-key-here"
      }
    }
  }
}
```

The API key can be obtained from LinkUp API service.

## Running

For development with auto-rebuild:
```bash
npm run watch
```

For production:
```bash
npm start
```

## Debugging

Since MCP servers communicate over stdio, debugging can be challenging. We recommend using the [MCP Inspector](https://github.com/modelcontextprotocol/inspector):

```bash
npm run inspector
```

## NEXT STEPS

Future improvements to consider:

1. Add caching for search results to improve performance
2. Implement pagination for large result sets
3. Add filtering options for search results
4. Support different output formats (markdown, HTML)
5. Add rate limiting and request throttling
6. Implement authentication for API access
7. Add more search parameters (date ranges, domains, etc.)
8. Improve error handling and user feedback
9. Add logging for debugging and monitoring
10. Implement health check endpoints

TDQS

A3.5/5.0

Scored across 1 tool

Disambiguation5/5

With only one tool, there is no possibility of confusion between tools. The tool's purpose is clearly stated as performing a deep web search.

Naming Consistency5/5

The single tool name 'deep_search' follows a clear verb_noun convention. There are no other tools to introduce naming inconsistencies.

Tool Count3/5

A single tool is borderline thin for a search server, which might benefit from additional operations like result filtering or history. However, for a very narrow scope, it could be acceptable.

Completeness3/5

The tool covers the core search action but lacks ancillary capabilities that users might expect, such as pagination, saved searches, or multi-phase queries. This leaves the surface minimal and somewhat incomplete for complex workflows.

Maintenance

ActivityInactive
ResponsivenessNo issues