Install the MCP server - Amazon Bedrock AgentCore
Add MCP server configurationVerify MCP server installation

Install the MCP server

To install the AgentCore MCP server, add it to your MCP client's configuration file. Each MCP client stores this configuration in a different location.

Add MCP server configuration

Choose your MCP client and add the corresponding configuration:

For Kiro

Add to .kiro/settings/mcp.json:


{
  "mcpServers": {
    "awslabs.amazon-bedrock-agentcore-mcp-server": {
      "command": "uvx",
      "args": ["awslabs.amazon-bedrock-agentcore-mcp-server@latest"],
      "env": {
        "FASTMCP_LOG_LEVEL": "ERROR"
      },
      "disabled": false,
      "autoApprove": []
    }
  }
}

For Cursor

Add to .cursor/mcp.json:


{
  "mcpServers": {
    "awslabs.amazon-bedrock-agentcore-mcp-server": {
      "command": "uvx",
      "args": ["awslabs.amazon-bedrock-agentcore-mcp-server@latest"],
      "env": {
        "FASTMCP_LOG_LEVEL": "ERROR"
      },
      "disabled": false,
      "autoApprove": []
    }
  }
}

For Amazon Q

The best practice is to configure MCP servers for individual Q CLI agents. For configuration instructions, see Configuring MCP servers for Amazon Q CLI.

For Amazon Q in IDEs (VS Code and JetBrains), see Using MCP servers with Amazon Q in your IDE.

For Claude Code

Configuration depends on your installation:

  • Standalone app: Add to ~/.claude/mcp.json

  • VS Code extension: Configure MCP servers through the Claude Code CLI first, then the extension will automatically use them. See the Claude Code VS Code documentation for setup details.

For standalone Claude Code app:


{
  "mcpServers": {
    "awslabs.amazon-bedrock-agentcore-mcp-server": {
      "command": "uvx",
      "args": ["awslabs.amazon-bedrock-agentcore-mcp-server@latest"],
      "env": {
        "FASTMCP_LOG_LEVEL": "ERROR"
      },
      "disabled": false,
      "autoApprove": []
    }
  }
}

Verify MCP server installation

To verify that the MCP server is connected and working successfully, restart your MCP client after adding the configuration and confirm that the following tools are available:

  • search_agentcore_docs - Search Amazon Bedrock AgentCore documentation

  • fetch_agentcore_doc - Fetch specific Amazon Bedrock AgentCore documentation pages