How to add a custom MCP server in CrewAI AMP
CrewAI's enterprise guide shows how to add a custom MCP server in AMP, including the URL, custom headers, and OAuth 2.0.
CrewAI's enterprise guide shows how to add a custom MCP server in AMP, including the URL, custom headers, and OAuth 2.0.
OpenHands docs show how to install the CLI with uv or the install script and start a first agent session.
Configure Google's remote Docs MCP server, enable the APIs, and authenticate so an agent can use Docs as a tool.
Set up the Claude Agent SDK, add an API key, and run a first agent that can read and edit files from the official quickstart.
Install the MCP Python SDK, register a typed tool, and open the server in the Inspector with mcp dev.
LangGraph's quickstart shows how to define tools, route tool calls, compile a graph, and invoke a first tool-using agent.
Vercel’s AI SDK docs show how to configure a ToolLoopAgent, cap its tool loop, and run it in a local terminal UI.
Official LangChain docs show how to pick a multi-agent pattern and call a LangChain agent from inside a custom LangGraph node.
Create a first stateful agent with the Letta Agent SDK: set an API key, install the package, open a session, and stream a reply.
Install the AgentCore CLI, scaffold a Python agent, test it locally, deploy to AWS, and invoke it from the terminal.
Install Nous Research Hermes Agent, choose a provider, start a terminal chat, and resume that session with the documented CLI.
Haystack’s get-started guide installs the packages and runs an Agent that calls a web-search tool through OpenAIChatGenerator.
LangChain’s quickstart installs the package and builds a tool-calling agent with create_agent, a model string, and agent.invoke.
The DeepSeek Harness Python SDK guide installs deepseek-harness-sdk and runs one task with an API key, a workspace, and a session id.
Call gemini-4-argon with the Interactions API, set a thinking level, and point a managed Antigravity agent at the same model.
MCP’s Agent Skills docs show how to install the mcp-server-dev plugin and let an agent scaffold a server after a short discovery pass.
Install the Strands Agents SDK and run a first in-process agent in Python or TypeScript against a model provider.
Create a first Mistral agent with the Agents API, save its ID, and optionally attach the built-in web search tool.
Install AG2, set a provider key, and run a first ConversableAgent that replies for up to three turns.
Create a Node project, register one weather tool with the MCP TypeScript SDK, and call it over stdio with the Inspector.
Message a Grok Bot, review drafts before send, and start a group chat of two to six Bots with explicit owners and limits.
Google’s Gemini API docs show how to call Gemini 3.8 Flash and the Antigravity preview agent through the Interactions API.
Create a Claude Managed Agent, attach an environment, and start a session with the official CLI or SDK quickstart.
Build LangGraph workflows that chain, parallelize, route, or spawn orchestrator workers with the Send API.
Install Hermes Agent from Nous Research on Linux, macOS, or Windows, then run hermes setup --portal.