Skip to main content
Aptly gives you two ways in. Use the MCP server to connect an AI assistant such as Claude or Codex with a simple sign-in. Use the Core API (REST) for scripts, integrations and more structured use cases, with fine-grained permissions on each API key. Not sure which? See the Introduction.

​
MCP quick start

Connect an AI assistant to Aptly in a couple of minutes. You only need the MCP server URL below. Aptly signs you in with OAuth, so there are no API keys to create or paste.
Permissions: the Aptly MCP server has the same broad permissions as the user who signs in. Anything that person can see and do in Aptly, the connected assistant can too, and it is not limited to particular boards or actions. Connect with an account whose access you are comfortable giving the assistant. If you need to restrict access, use the Core API with a scoped API key instead.
1

Open Connectors

In Claude Desktop, open Settings and select Connectors.
2

Add a custom connector

Click Add custom connector. Enter a name such as Aptly and paste the MCP server URL:
Click Add.
3

Sign in to Aptly

Click Connect. A browser window opens. Sign in to Aptly and approve access.
4

Try it

Start a new chat, make sure the Aptly connector is turned on, and ask something like “Search my Aptly boards for Maintenance” or “Show my latest unread emails”.
On Claude Team and Enterprise plans, an organization Owner may need to add the connector for the organization first. Members then find it under Connectors and click Connect.
See MCP tools for the full list of what the assistant can do.

​
Core API quick start

Use the REST API directly from scripts and integrations with an API key. Set each key’s permissions under Setup → Developer in Aptly.
1

Enable the API for your board

In Aptly, open the board → Card Sources → API and toggle the API on.
2

Create an API key

Under the API section, click Create New Key, give it a name, and copy the key value.
3

Fetch the schema

This returns the list of field keys you’ll need for reading and writing cards.
4

Create a card

​
Using Rest API with Claude/Codex

Coding assistants such as Claude Code and Codex can call the Core API for you from your computer. You give them your API key in a local .env file and a short instruction file that tells them how to use it.
This works with tools that can run commands on your machine: Claude Code (the terminal app, or the Code tab in Claude Desktop) and Codex (desktop app, CLI or IDE extension). The regular chat in Claude Desktop can’t read local files, so use the MCP server there instead.
1

Create a scoped API key

In Aptly, create an API key under Setup → Developer and give it only the access the assistant needs (boards, inboxes, and read, insert or update). See Core API Base URL for the options. Copy the key.
2

Save the key in a .env file

In the folder you open in Claude Code or Codex, create a file named .env:
Then make sure the file is never committed to git. If the folder is a git repository, add it to .gitignore:
3

Tell the assistant how to use it

Add the instructions below to an instruction file in the same folder. Claude Code reads CLAUDE.md and Codex reads AGENTS.md. Create whichever you use (or both).
4

Ask for something

Open the folder in your tool and ask, for example, “List my Aptly boards” or “Show the fields on my Maintenance board”. The first time, the tool will ask your permission to run the request.
  • Open the folder with claude in your terminal, or choose it in the Code tab of Claude Desktop.
  • Claude Code asks before running each command. When it asks to run a read request such as listing boards, you can choose to always allow that kind of command. Keep requests that create, change or delete data on “ask”.
Anything the assistant can read, it can send to its AI provider as part of the conversation. Don’t ask it to open or print your .env file. Use a key with the least access needed, and revoke the key in Setup → Developer if it is ever exposed.