Agentup
Portable , scalable , secure AI Agents
What is Agentup?
Agentup is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to portable , scalable , secure ai agents
Portable , scalable , secure AI Agents
This server falls under the Coding Agents category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- Portable , scalable , secure AI Agents
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx agentupConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use Agentup
AgentUp is a Python framework for building portable, scalable, and secure AI agents that communicate using the Agent-to-Agent (A2A) protocol and expose their capabilities through MCP. Agents are defined through a declarative configuration file rather than boilerplate code, and the framework handles authentication (OAuth2, JWT, API keys), task routing, async message-driven execution, and automatic Agent Card generation for inter-agent discovery. Teams use AgentUp to deploy production-ready multi-agent systems where individual agents can be independently versioned, secured, and composed.
Prerequisites
- Python 3.10+ with pip
- An MCP-compatible client such as Claude Desktop (for connecting to AgentUp-powered MCP tools)
- Optional: Docker and Redis for production deployments with async task queues
- Optional: an AI provider API key (OpenAI, Anthropic, etc.) for the agent's underlying LLM
Install AgentUp
Install the AgentUp framework from PyPI using pip.
pip install agentupInitialise a new agent project
Run `agentup init` to launch an interactive configuration wizard. It will prompt you for the agent name, capabilities, authentication method, AI provider, and transport settings.
mkdir my-agent && cd my-agent
agentup init
# Follow the interactive prompts to configure:
# - Agent name and description
# - Capabilities (MCP tools, plugins)
# - Authentication type (OAuth2, JWT, API key, or none)
# - AI provider and modelReview the generated configuration file
AgentUp generates a declarative configuration file. Review it to add or adjust capabilities, authentication scopes, and MCP tool definitions before running.
# The configuration file defines your entire agent declaratively.
# Edit it to add capabilities, plugins, or adjust auth settings.
cat agentup.yamlStart the agent
Run `agentup run` to start the agent. It launches on localhost:8000 by default with a JSON-RPC API, MCP tool endpoints, and the configured authentication middleware.
agentup run
# Agent starts at http://localhost:8000
# MCP endpoint available at http://localhost:8000/mcp
# Agent Card at http://localhost:8000/.well-known/agent.jsonConnect a client or integrate with other agents
Connect Claude Desktop or another MCP client to the running agent's MCP endpoint. Other AgentUp agents can discover this agent via its auto-generated Agent Card.
# For Claude Desktop, add the running agent as an MCP server:
# {
# "mcpServers": {
# "my-agent": {
# "url": "http://localhost:8000/mcp"
# }
# }
# }
# Or use AgentUpChat for a built-in chat UI:
# https://github.com/RedDotRocket/AgentUpChatAgentup Examples
Client configuration (Claude Desktop — HTTP MCP endpoint)
Connect Claude Desktop to a running AgentUp agent that exposes an MCP endpoint over HTTP. Start the agent with `agentup run` first.
{
"mcpServers": {
"agentup": {
"url": "http://localhost:8000/mcp"
}
}
}Prompts to try
These prompts depend on the capabilities configured in your agentup.yaml. The framework supports both reactive (single-shot) and iterative (multi-step) agent types.
- "What capabilities does this agent have?"
- "Execute the research task: summarise the latest news about AI regulation."
- "List all registered agents and their capabilities."
- "Delegate this task to the most appropriate sub-agent: analyse sales data from last quarter."
- "Run a multi-step workflow: gather data, analyse trends, and produce a report."Troubleshooting Agentup
`agentup run` fails with a missing configuration error
Ensure you have run `agentup init` in the current directory and that the agentup.yaml configuration file exists. Run `agentup init` again if the file is missing or corrupted.
MCP client cannot connect to the agent on port 8000
Confirm the agent is running by visiting http://localhost:8000/.well-known/agent.json in your browser. If a firewall is blocking connections, ensure port 8000 is open. To use a different port, set the port in your agentup.yaml configuration before running.
Authentication errors when calling agent tools
If you configured OAuth2 or JWT authentication during `agentup init`, your MCP client must include valid credentials. For local development, use the 'none' or 'api_key' authentication type to avoid auth complexity. Set the correct API key in your MCP client's env block.
Frequently Asked Questions about Agentup
What is Agentup?
Agentup is a Model Context Protocol (MCP) server that portable , scalable , secure ai agents It connects AI assistants to external tools and data sources through a standardized interface.
How do I install Agentup?
Follow the installation instructions on the Agentup GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with Agentup?
Agentup works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is Agentup free to use?
Yes, Agentup is open source and available under the Apache-2.0 license. You can use it freely in both personal and commercial projects.
Agentup Alternatives — Similar Coding Agents Servers
Looking for alternatives to Agentup? Here are other popular coding agents servers you can use with Claude, Cursor, and VS Code.
Dify
★ 142.2kProduction-ready platform for agentic workflow development.
Ruflo
★ 54.0k🌊 The leading agent orchestration platform for Claude. Deploy intelligent multi-agent swarms, coordinate autonomous workflows, and build conversational AI systems. Features enterprise-grade architecture, self-learning swarm intelligence, RAG integrat
Goose
★ 45.7kan open source, extensible AI agent that goes beyond code suggestions - install, execute, edit, and test with any LLM
Antigravity Awesome Skills
★ 38.3kInstallable GitHub library of 1,400+ agentic skills for Claude Code, Cursor, Codex CLI, Gemini CLI, Antigravity, and more. Includes installer CLI, bundles, workflows, and official/community skill collections.
AgentScope
★ 25.5kBuild and run agents you can see, understand and trust.
Serena
★ 24.5kA coding agent toolkit that provides IDE-like semantic code retrieval and editing tools, enabling LLMs to efficiently navigate and modify codebases using symbol-level operations instead of basic file reading and string replacements.
Browse More Coding Agents MCP Servers
Explore all coding agents servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up Agentup in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use Agentup?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.