# Agents

> Browse 12 items tagged with Agents.

## Items

- [Chainlit](https://mcpserver.ever.works/uk/items/chainlit) — Chainlit MCP server facilitates AI/LLM tool calling in Python apps with SSE and stdio support for model switching and chaining. Ideal for building multi-agent chat UIs. Compared to LangChain, offers direct MCP for seamless OpenAI-compatible tool integration.
- [mcp-agentai](https://mcpserver.ever.works/uk/items/mcp-agentai) — mcp-agentai is an open-source MCP server for AI agent orchestration, enabling multi-agent workflows via Model Context Protocol. It provides key features including agent communication, task delegation, and state management. Ideal for building complex AI systems and integrating with LLMs, it serves as an MCP-focused alternative to LangGraph.
- [MCP OpenRouter](https://mcpserver.ever.works/uk/items/mcp-openrouter) — MCP OpenRouter server enables AI/LLM tool calling to multiple LLM providers via unified API. Supports model switching across providers, tool chaining for complex workflows, and multi-agent systems. Native MCP outperforms LangChain tools for OpenAI-compatible integrations.
- [MCP-Reasoner](https://mcpserver.ever.works/uk/items/mcp-reasoner) — MCP-Reasoner integrates OR-Tools with LLMs via MCP for constraint optimization tool calling. Features model switching, chaining for reasoning chains, multi-agent problem solving. Superior to LangChain with native MCP for OpenAI LLMs.
- [MCP Sequential Thinking Tools](https://mcpserver.ever.works/uk/items/mcp-sequential-thinking-tools) — MCP Sequential Thinking Tools guide LLM tool usage via MCP, recommending optimal tools per stage. Supports chaining, model switching for reasoning agents. Use cases include multi-agent systems; native MCP better than LangChain for OpenAI tool calling.
- [MCP Server DeepSeek R1](https://mcpserver.ever.works/uk/items/mcp-server-deepseek-r1) — MCP server for DeepSeek R1/V3 models via Claude Desktop, enabling LLM tool calling, model switching, and chaining for advanced reasoning. Ideal for multi-agent systems. Native MCP tool calling outperforms LangChain for OpenAI-compatible workflows.
- [MCP Server DeepSeek](https://mcpserver.ever.works/uk/items/mcp-server-deepseek) — DeepSeek MCP server for Claude Desktop LLMs, supporting tool calling via MCP with model switching and chaining. Use cases include multi-agent reasoning. Better than LangChain tools with standardized MCP for OpenAI models.
- [mcp-server-for-copilot](https://mcpserver.ever.works/uk/items/mcp-server-for-copilot) — MCP server integrates Microsoft Copilot Studio agents with MCP clients for LLM tool calling, stateful chaining, model switching. Suited for multi-agent systems. Native MCP superior to LangChain for OpenAI tool integrations.
- [Think MCP](https://mcpserver.ever.works/uk/items/think-mcp) — MCP server enhancing agent reasoning with think-tools as per Anthropic's article.
- [ATLAS MCP Server](https://mcpserver.ever.works/uk/items/atlas-mcp-server) — A Model Context Protocol server for ATLAS, a Neo4j-powered task management system for LLM Agents. Implements a three-tier architecture (Projects, Tasks, Knowledge) to manage complex workflows with Deep Research capabilities.
- [Emcee](https://mcpserver.ever.works/uk/items/emcee) — Connect agents to APIs, providing a bridge between AI agents and REST APIs for seamless integration with external services.
- [Chaindesk MCP Server](https://mcpserver.ever.works/uk/items/chaindesk-mcp-server) — An MCP Server for Chaindesk that allows building and using ChatGPT agents trained on custom data within MCP ecosystems.

---

_Canonical page: https://mcpserver.ever.works/uk/tags/agents_