Fast Agent
Code, Build and Evaluate agents - excellent Model and Skills/MCP/ACP Support
What is Fast Agent?
Fast Agent is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to code, build and evaluate agents - excellent model and skills/mcp/acp support
Code, Build and Evaluate agents - excellent Model and Skills/MCP/ACP Support
This server falls under the Coding Agents category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- Code, Build and Evaluate agents - excellent Model and Skills
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx fast-agentConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use Fast Agent
fast-agent is a Python framework and MCP-compatible agent runtime for building, composing, and evaluating AI agents with rich model and skill support. It lets you define agents with simple decorators, chain them into workflows (sequences, parallel fan-out, routers, evaluator-optimizers), and connect them to any MCP server over stdio, HTTP, or SSE transport. Developers use it to prototype and ship multi-agent systems that work with Anthropic, OpenAI, Google, Azure, Ollama, and other model providers without changing application code.
Prerequisites
- Python 3.10 or later installed
- uv package manager installed (https://github.com/astral-sh/uv)
- API key for your chosen model provider (e.g. Anthropic, OpenAI, Google)
- An MCP client such as Claude Desktop, or use fast-agent's own interactive CLI
- Node.js (optional, for connecting to Node-based MCP servers)
Install fast-agent-mcp
Install the fast-agent-mcp package globally using the uv tool manager, or run it directly with uvx for a one-off execution.
uv tool install -U fast-agent-mcpCreate a fast-agent.secrets.yaml file
Store your model provider API keys in a secrets file in your project directory. This file is separate from the main config so you can safely version-control fast-agent.yaml without leaking credentials.
# fast-agent.secrets.yaml
anthropicApiKey: "sk-ant-..."
openaiApiKey: "sk-..."Create a fast-agent.yaml configuration file
Define your MCP servers, model aliases, and other settings in fast-agent.yaml placed in your project root.
# fast-agent.yaml
mcp:
servers:
everything:
transport: stdio
command: npx
args: ["-y", "@modelcontextprotocol/server-everything"]
default_model: claude-sonnet-4-6Write a basic agent
Create a Python file defining your agent using the @fast.agent decorator. Agents can be interactive (waiting for user input) or run programmatically.
import asyncio
from fast_agent import FastAgent
fast = FastAgent("My Agent")
@fast.agent(instruction="You are a helpful assistant.")
async def main():
async with fast.run() as agent:
await agent.interactive()
if __name__ == "__main__":
asyncio.run(main())Run the agent
Execute your agent file with uv run, optionally overriding the model at the command line.
uv run agent.py --model=claude-sonnet-4-6Launch a pre-built agent pack
Use the fast-agent CLI to run a pre-configured agent pack directly without writing code. The --pack flag downloads and runs a community-built agent bundle.
uvx fast-agent-mcp@latest --pack analyst --model haikuFast Agent Examples
Claude Desktop configuration
To expose a fast-agent as an MCP server, launch it with HTTP transport. Connect Claude Desktop to the running server using the streamable_http transport.
{
"mcpServers": {
"fast-agent": {
"command": "uv",
"args": ["run", "agent.py", "--transport", "http", "--port", "8001"],
"env": {}
}
}
}Prompts to try
Use these interactive commands inside a running fast-agent session to manage connections and skills.
- "Fetch the latest headlines from https://hf.co/mcp and summarize them"
- "Use the everything server tool to list available resources"
- "/connect @modelcontextprotocol/server-filesystem" (attach a new MCP server mid-session)
- "/skills" (list available agent skills)
- "Chain a URL fetcher and social media writer to create a tweet thread from http://example.com"Troubleshooting Fast Agent
Agent fails to connect to an MCP server defined in fast-agent.yaml
Check that the command and args for the MCP server are correct and that the executable is on your PATH. Use the MCP Ping utility (fast-agent go --url <server-url>) to verify transport connectivity before wiring it into an agent.
Model provider authentication errors at runtime
Ensure fast-agent.secrets.yaml exists in the same directory where you run the agent, and that the key names match expected fields (anthropicApiKey, openaiApiKey, etc.). The secrets file takes priority over environment variables.
uvx command not found or fast-agent-mcp not resolving
Install uv first with: curl -LsSf https://astral.sh/uv/install.sh | sh. Then run: uv tool install -U fast-agent-mcp. Verify with: fast-agent --version
Frequently Asked Questions about Fast Agent
What is Fast Agent?
Fast Agent is a Model Context Protocol (MCP) server that code, build and evaluate agents - excellent model and skills/mcp/acp support It connects AI assistants to external tools and data sources through a standardized interface.
How do I install Fast Agent?
Follow the installation instructions on the Fast Agent GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with Fast Agent?
Fast Agent works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is Fast Agent free to use?
Yes, Fast Agent is open source and available under the Apache-2.0 license. You can use it freely in both personal and commercial projects.
Fast Agent Alternatives — Similar Coding Agents Servers
Looking for alternatives to Fast Agent? 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 Fast Agent 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 Fast Agent?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.