My experiments with MCP (Model Context Protocol)

Lightweight framework standardizing storage and communication of training config, data lineage, hyperparameters, environment, and deployment metadata in a single archive.

RailTel Model Context Protocol Model

Distributed orchestration delivering integrated product discovery, price comparison, and sentiment-based reviews powered by IBM watsonx.

A full-stack demo showcasing a model-context workflow with SSE transport, Claude integration, and tool calls for guitar recommendations.

Modular MCP server enabling code-aware retrieval and automated README generation with LLM-powered documentation.

Model Context Protocol

Enables Claude AI to write and execute code within Jupyter Notebooks.

Lightweight protocol enabling modular context handling and dynamic tool integration for agentic AI.

Lightweight sentiment analysis backend with STDIO transport and MCP protocol integration, featuring Inspector-driven debugging for rapid local experimentation.

Integrates Google Gemini 2.5 Pro with SerpAPI to enable real-time web search, contextual reasoning, and seamless tool integration for enhanced AI workflows.

LLM Model Context Protocol (MCP) demo using Spring AI

FastAPI-based context processor linking a local Llama model with GitHub repo file listing to enrich queries.

ReAct-style agent demonstrating standardized integration between LLMs and external tools via a universal connector.

Autonomous AI agent system delivering dynamic context retrieval, task decomposition, and secure collaboration to accelerate development workflows.

Provides programmatic AlphaFold predictions, UniProt summaries, and protein annotations via FastMCP for rapid research integration.

Model Context Protocol Example

Standardized context sharing enables coordinated multi-agent workflows with dynamic routing and evolving prompts.

Protocol-driven framework enabling seamless model-context integration, dynamic context shaping, cross-model coordination, and extensible plugin ecosystems.

Model Context Protocol

Open protocol unifying LLMs with diverse data sources and tools via streamlined middleware.

Model Context Protocol (MCP) is an open protocol that standardizes how applications provide context to LLMs- https://docs.anthropic.com/en/docs/agents-and-tools/mcp

Prototype illustrating the Model Context Protocol workflow, explaining how Anthropic MCP operates through a lightweight Python agent. It demonstrates context handling, prompt structuring, and basic tool interactions, offering a clear, educational reference for developers exploring MCP concepts and integration patterns. Minimal by design to highlight core interactions.

Arxiv knowledge agent blends Arxiv search with modular tool integration and a user-friendly web/CLI interface for efficient research.