🦅 Saqr-MCP

What is 🦅 Saqr-MCP?

🦅 Saqr-MCP is a Python application that implements the Model Context Protocol (MCP) to enable advanced AI assistant capabilities. It supports both local models through Ollama and cloud models through Groq, providing a flexible client-server architecture with tools for web search, memory management, document generation, and advanced reasoning.

How to use 🦅 Saqr-MCP?

Install dependencies with UV, configure environment variables (API keys for Tavily, Groq, Mem0; Ollama model name), and run python main.py. Type queries in the interactive console; use quit to exit. To use Groq instead of Ollama, modify main.py to import from src.groq_client.

Key features of 🦅 Saqr-MCP

• Interactive chat interface for querying models
• Support for local models (Ollama) and cloud models (Groq)
• Advanced web search capabilities using Tavily API
• Word document generation from markdown content
• Comprehensive memory management system using mem0
• Advanced reasoning and thought process tracking

Use cases of 🦅 Saqr-MCP

• Perform real-time web searches to retrieve up-to-date information
• Create and manage persistent memories for context-aware interactions
• Generate formatted Word documents from markdown content
• Record and analyze reasoning processes during complex problem-solving
• Switch between local and cloud models depending on availability or cost

FAQ from 🦅 Saqr-MCP

What are the prerequisites for using 🦅 Saqr-MCP?

Python 3.11 or higher, Ollama installed for local model usage, and the UV package manager (recommended).

What API keys are required and where do I get them?

You need a Tavily API key from app.tavily.com, a Groq API key from console.groq.com, and a Mem0 API key from mem0.ai. Configure them in a .env file based on .env.example.

How do I switch between local and cloud models?

By default the app uses Ollama. To use Groq, edit main.py to import SaqrMCPClient from src.groq_client instead of src.ollama_client.

What tools are available on the server?

Tools include web_search, word_file_generator, add_memory, get_all_memories, search_memories, think, get_thoughts, clear_thoughts, and get_thought_stats.

What is the project structure?

The entry point is main.py. The src/ folder contains ollama_client.py, groq_client.py, server.py, and logger.py.