MCP Setup

Connect Chronary to your MCP client. You’ll need an API key from the Chronary console before starting.

Two transports — pick whichever your host supports

This page covers the hosted HTTP/SSE transport at https://api.chronary.ai/mcp — recommended for Claude Desktop, Cursor, VS Code Copilot, and any host that speaks streamable HTTP MCP. No install step.

If your host only speaks stdio, or you want to pin a specific server version, run the @chronary/mcp stdio server locally via npx -y @chronary/mcp. Same 54 tools, same API key — just a different transport. The package README has per-editor config snippets (incl. the Windows cmd /c wrapper).

Install via Smithery (one-click)

Chronary is listed on Smithery, the MCP registry — add it to any supported client without hand-editing config files:

• From the web: open the Chronary listing and click Add to toolbox, then choose your client.
• From the terminal:

  npx -y @smithery/cli@latest mcp add chronary/agent-calendar --client claude

  Swap --client for cursor, vscode, claude-code, windsurf, cline, goose, and others. You’ll be prompted for your Chronary API key.

Prefer to wire it up yourself? Use the manual per-client config below.

Client setup

Claude Desktop

Add Chronary to your Claude Desktop configuration file:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "chronary": {
      "url": "https://api.chronary.ai/mcp",
      "headers": {
        "Authorization": "Bearer chr_sk_your_key_here"
      }
    }
  }
}

Restart Claude Desktop after saving. You should see a hammer icon indicating MCP tools are available.

Cursor

Open Cursor Settings and navigate to the MCP section. Add a new server with these settings:

• Name: chronary
• Transport: Streamable HTTP
• URL: https://api.chronary.ai/mcp

In your Cursor MCP configuration file (.cursor/mcp.json in your project root or global config):

{
  "mcpServers": {
    "chronary": {
      "url": "https://api.chronary.ai/mcp",
      "headers": {
        "Authorization": "Bearer chr_sk_your_key_here"
      }
    }
  }
}

Restart Cursor after saving. Chronary tools will appear in the agent tool list.

TypeScript

Install the MCP SDK:

npm install @modelcontextprotocol/sdk

Connect to Chronary:

import { Client } from '@modelcontextprotocol/sdk/client/index.js';
import { StreamableHTTPClientTransport } from '@modelcontextprotocol/sdk/client/streamableHttp.js';

const client = new Client({
  name: 'my-agent',
  version: '1.0.0',
});

const transport = new StreamableHTTPClientTransport(
  new URL('https://api.chronary.ai/mcp'),
  {
    requestInit: {
      headers: {
        Authorization: 'Bearer chr_sk_your_key_here',
      },
    },
  }
);

await client.connect(transport);

// List available tools
const { tools } = await client.listTools();
console.log(`Connected — ${tools.length} tools available`);

// Call a tool
const result = await client.callTool({
  name: 'list_agents',
  arguments: {},
});
console.log(result);

Note

The @modelcontextprotocol/sdk package supports Node.js 18+ and all modern JavaScript runtimes (Deno, Bun, and edge-runtime hosts).

Python

Install the MCP package:

pip install mcp

Connect to Chronary:

from mcp.client.streamable_http import streamablehttp_client
from mcp import ClientSession

async def main():
    async with streamablehttp_client(
        url="https://api.chronary.ai/mcp",
        headers={"Authorization": "Bearer chr_sk_your_key_here"},
    ) as (read_stream, write_stream, _):
        async with ClientSession(read_stream, write_stream) as session:
            await session.initialize()

            # List available tools
            tools = await session.list_tools()
            print(f"Connected — {len(tools.tools)} tools available")

            # Call a tool
            result = await session.call_tool("list_agents", arguments={})
            print(result)

import asyncio
asyncio.run(main())

Note

The mcp package requires Python 3.10+. It uses httpx under the hood for async HTTP.

Verify your connection

After setup, test that everything is working. Ask your AI client:

“List my agents”

The AI should call the list_agents tool and return results. If you haven’t registered any agents yet, you’ll see an empty list — that’s expected.

Confirm tools are visible

In Claude Desktop, click the hammer icon to see all available tools. You should see 50 Chronary tools listed.

In Cursor, check the agent tool list in the MCP settings panel.

For custom agents, call client.listTools() (TypeScript) or session.list_tools() (Python) and verify 54 tools are returned.

Register a test agent

Ask: “Register an AI agent called Test Agent”

The AI should call create_agent and return an agent with an agt_ prefixed ID.

Create a calendar and event

Ask: “Create a calendar called ‘Demo Calendar’ in UTC for that agent, then add a meeting tomorrow at 2pm”

The AI should call create_calendar followed by create_event.

Verify via REST (optional)

Confirm the data is shared by querying the REST API:

curl https://api.chronary.ai/v1/agents \
  -H "Authorization: Bearer chr_sk_your_key_here"

You should see the same agent and calendar created via MCP.

Troubleshooting

Problem	Solution
Tools not appearing	Restart your client after editing the config file. Check for JSON syntax errors.
401 Unauthorized	Verify your API key is correct and includes the Bearer prefix.
403 Forbidden	Your API key may be deactivated. Check the console.
Connection timeout	Ensure you can reach api.chronary.ai from your network. Check firewall/proxy settings.
Tool call returns error	Check the error message — it follows the same error format as the REST API.

Next steps

Tools ReferenceAll 50 tools with parameters and examples

MCP OverviewArchitecture, transport, and how MCP relates to REST

AuthenticationTest vs live keys, key rotation, and scopes