Atlassian MCP server: ask questions, get governed answers

Atlassian dashboard questions you can ask

Typical questions from engineering and project teams. Each answer shows the definition, source and query behind it.

• How many open bugs are there by project and priority?
• What was our sprint velocity over the last six sprints?
• Which issues have been in progress for more than 14 days?
• Show issues created versus resolved per week this quarter.
• Which assignees have the most unresolved issues right now?
• Which Confluence spaces had the most pages updated in the last 30 days?

Worked example: a governed Jira & Confluence answer

One question, and everything the answer shows so you can check it. The figures are invented sample data.

project = PAY
  AND sprint in ("PAY Sprint 41", "PAY Sprint 42", "PAY Sprint 43")
  AND statusCategory = Done
ORDER BY sprint ASC

Common pitfalls when asking Jira & Confluence questions

Each one is settled once in an approved definition, so every person and every assistant gets the same answer.

• Story points live in different fields

  Company-managed and team-managed projects store estimates in different fields. The definition names the field per project, so a cross-project velocity is not silently zero for some teams.

• Done is a status category, not a status name

  Workflows call the final step Closed, Resolved or Released. Counting by status category catches all of them; counting by one status name misses the rest.

• Issues carried over between sprints

  An unfinished issue appears in every sprint it was part of. Velocity counts it once, in the sprint it was completed, or carried-over work inflates the trend.

How Nexatron connects to Atlassian

Whose permissions

Each person connects their own Atlassian account, so answers return only the records that person can already see in Atlassian.

Read and write

Questions are read-only. A small set of Atlassian actions also exists; each one is previewed first, runs only after explicit confirmation by a permitted person, and can be switched off by your administrator.

Setup

Sign-in happens on the vendor's own OAuth consent screen; Nexatron never sees a password. Steps: choose the source → check prerequisites → enter connection details → approve access → verify → ask. No data migration and no pipeline to build: data is queried where it lives.

Atlassian MCP connector for Claude, ChatGPT and Copilot

Nexatron's hosted MCP server is a governed Atlassian MCP connector: Claude, ChatGPT and Microsoft Copilot ask questions of your Atlassian data through it. Each person signs in as themselves, so the assistant sees only what they may see.

Connect Atlassian to Claude, ChatGPT or Copilot

1. Connect Atlassian to your Nexatron workspace.
2. In Claude, add Nexatron from Claude's connector directory. In ChatGPT or Copilot, add a custom MCP connector with the server URL https://api.nexatron.io/api/v1/mcp.
3. Sign in with your own Nexatron identity (OAuth 2.1), then ask your Atlassian questions.

Atlassian connector FAQ

Can I ask Atlassian questions without writing queries?

Yes. The connector is available now. Ask in Teams, Slack, the Nexatron web app, or an AI assistant connected through Nexatron's MCP server.

Does Nexatron respect Atlassian permissions?

Yes. With a personal connection, answers return only the Atlassian records that person can already see.

Can Nexatron change data in Atlassian?

Only through optional actions that a permitted person previews and confirms. Asking a question never changes Atlassian data.

What do I need to connect Atlassian?

A user who can approve access on the vendor's own sign-in screen.

Is there a Atlassian MCP server for Claude, ChatGPT or Copilot?

Yes: Nexatron's hosted MCP server, once Atlassian is connected to your workspace. It is Nexatron's server, not one published by Atlassian, and the assistant gets the same governed answer as the Nexatron app.

Can I get Jira dashboard answers without building a new dashboard?

Yes. Ask the question and Nexatron returns the number with the JQL and definition behind it. Jira dashboards stay useful for the views your team watches daily; Nexatron covers the one-off and follow-up questions in between.

Can I ask about Confluence pages as well as Jira issues?

Yes. The same Atlassian connection covers Jira issues, sprints and projects, and page activity in Confluence spaces, such as which spaces had the most pages updated in the last 30 days.

How is Nexatron different from Atlassian's own AI features?

Atlassian's AI features work inside Atlassian products. Nexatron answers from Teams, Slack, the Nexatron web app or a connected assistant, can combine Jira with other connected systems when federation is on (the default for new workspaces), and shows the definition and query behind each number.