> ## Documentation Index
> Fetch the complete documentation index at: https://www.plain.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Connect Jira

> Connect Sidekick to Jira through the Atlassian Rovo MCP server, using a service account instead of a personal account.

Sidekick's Jira tool connects through Atlassian's Rovo Model Context Protocol (MCP) server, authenticated with a service account instead of a personal login. Atlassian defines a service account as an account that is not associated with a person, so the connection keeps working if the person who set it up leaves your team.

## Before you start

You need:

* **Org admin** access to Atlassian Administration (`admin.atlassian.com`), to create the service account
* Workspace **Admin** or **Owner** access in Plain, to connect the server in Sidekick's settings

## Set up a service account in Atlassian

Do this in Atlassian Administration, before touching Sidekick's settings.

<Steps>
  <Step title="Enable API token authentication">
    Go to **Atlassian Administration → Rovo → Rovo MCP server**. In the **Authentication** section, turn **API token** on.

    This is separate from creating the credential itself. Both steps are required.
  </Step>

  <Step title="Create or reuse a service account with Jira access">
    Go to **Directory → Service accounts**.

    * **New service account**: click **Add service account**, name it something identifiable such as `Sidekick MCP`, and select **Jira** as one of its apps.
    * **Existing service account**: open it, select **Grant access**, and select **Jira**.
  </Step>

  <Step title="Create an API key">
    On the service account's page, select **Create credentials → API token**.

    Select these scopes:

    | Scope             | Why Sidekick needs it                                                    |
    | ----------------- | ------------------------------------------------------------------------ |
    | `read:me`         | Resolves the connected identity. Required before any other call succeeds |
    | `read:account`    | Resolves the connected identity. Required before any other call succeeds |
    | `read:jira-work`  | Reads and searches issues                                                |
    | `write:jira-work` | Creates and edits issues and comments                                    |

    Set an expiration and copy the API key. Atlassian shows it once.
  </Step>
</Steps>

## Connect the server in Sidekick

With the service account and API key ready, add the server as a custom MCP server in Sidekick.

<Steps>
  <Step title="Open the Add MCP server form">
    Go to **Plain AI → Sidekick → Integrations**, then in **Custom MCP servers** click **Add MCP server**.
  </Step>

  <Step title="Fill in the server details">
    * **Name**: `Atlassian`
    * **Server URL**: `https://mcp.atlassian.com/v2/mcp`
    * **Authentication**: API key
  </Step>

  <Step title="Add the Authorization header">
    Under **Auth parameters**, set:

    * **Placement**: Header
    * **Name**: `Authorization`
    * **Value**: `Bearer` followed by a space and the API key you copied, for example `Bearer your_atlassian_service_account_api_key`
  </Step>

  <Step title="Click Connect">
    Sidekick opens the connection and lists the tools Atlassian's server makes available.
  </Step>
</Steps>

## Verify the connection

A successful connection only confirms Sidekick can list Atlassian's tools, not that it can read your Jira data. Confirm both in a Sidekick chat: ask it to search Jira for issues in one of your real projects. Watch the tool call render in the chat. If it returns real issues, the connection works end to end.