Import from Jira

When can I import from Jira?

When you create a project. A company admin presses Create project, sets the name, key and board type, and under Start with picks Import from Jira instead of Empty project. The import has six steps: Project, Source, Columns, People & types, Review and Import.

• Use the same key as in Jira and issue numbers stay identical (NCP-123 stays NCP-123)
• Only a project created with Import from Jira can import — a project started empty has no import at all
• Works for Scrum, Kanban and Delivery pipeline boards

What does connecting to Jira Cloud bring over?

The recommended way: enter your Jira site (your-company.atlassian.net), your Jira account email and a scoped, read-only API token created at id.atlassian.com → Security → API tokens, then pick one of the Jira projects that account can see.

• Everything: issues, descriptions, comments with their authors, status history (for a backdated burndown), sprints with dates and goals, parents and sub-tasks, issue links, attachments and inline images
• Column order comes from your Jira board
• The token is used for this import only and never saved, and only your Jira site is ever called
• Scopes to select when you create the token (all read-only): read:jira-work, read:jira-user, read:board-scope:jira-software, read:board-scope.admin:jira-software, read:epic:jira-software, read:issue:jira-software, read:sprint:jira-software and read:project:jira
• Jira Cloud only (*.atlassian.net)

Can I import from a CSV file instead?

Yes — the backup for Jira Data Center, or when you no longer have Jira access: in Jira use Filters → Export → Export Excel CSV (all fields). Jira exports 1,000 issues per file, so add every file in one import (up to 30 files, 20 MB each, 20,000 issues); an issue key found in two files is imported once.

• CSV brings data only: no attachments or images (their file names and original Jira links are kept as text at the end of each issue), no sprint dates and no status history — the burndown starts from the import date
• Pick the date format in the file: detect automatically, English, Thai month names or ISO — it depends on the exporter's Jira profile language
• Rows that can't be read are skipped and listed on the Review step

Where do Jira statuses go on the board?

The Columns step lists every Jira status with its issue count and Jira category. Suggestions are filled in; change any row and check the board preview, where you can reorder columns.

• Same name as a column (any case) → merged into that column
• No matching column → Backlog with the label jira-status:<name>, except statuses Jira counts as done (such as Closed or Resolved) → Done with the same label
• Change a row to merge into an existing column or to create a new column with the Jira name (not on Delivery pipeline, whose columns follow its environments)
• A status named On hold goes to the On Hold column after Backlog (Kanban and Delivery pipeline), and any status can pick the On Hold column
• A Delivery pipeline matches environment names for you, such as UAT DONE → Passed (UAT), Testing UAT → Testing (UAT) and In Progress → Developing
• Kanban and Scrum: Same columns as Jira builds one column per Jira status in the order of your Jira board in one click (empty columns included when the token has the read:project:jira scope)
• Default columns no Jira status goes to (such as In Review or Test Rejected) can be left out: × on the column in the preview, or Only columns with work, so the board looks like Jira; put them back any time (Kanban keeps Backlog, and one Done column always stays)
• On Scrum, “Backlog” means To Do and not in a sprint
• At most 32 columns, at least one in the Done category, no duplicate names, and Kanban starts with the Backlog column

How are people, issue types and priorities imported?

• Jira doesn't share email addresses, so people are matched by name — change any match; people left unmatched become unassigned and keep their Jira name on the issue and their comments (invite them first, or reassign later)
• Jira types map to Epic, Task, Subtask or Bug (Story, Spike and others → Task, Sub-task → Subtask), and the original type is kept as a label such as jira-type:Story
• Jira priorities go into the Priority field: Highest, High, Medium, Low and Lowest match as they are, the old Jira Server names (Blocker, Critical, Major, Minor, Trivial) become Highest to Lowest in that order, and for any other name you pick the level on the People & types step (Medium is suggested) and it also gets a priority:<name> label so nothing is lost; issues without a priority become Medium
• Custom fields are added to the description as a table, and created/updated dates come from Jira

How do I review and run the import?

Review sums up the issues, comments, sprints, new columns and attachments with their size against your company's storage; if the attachments would pass the storage ceiling or the free tier, the import can't start until that is fixed. Then press Create project & import.

• The project is read-only while importing; the screen shows the time elapsed, the count of the current step (such as attachments 860 / 2,736) and about how long it still takes. You can close the window — you are notified when it finishes or fails
• If it fails, nothing is kept: the project is empty and can be imported again
• When done, download the import report (.md): skipped rows, attachments that couldn't be copied (kept as links back to Jira) and unmatched people
• History records it as “import from Jira” commits by the person who ran it

Can I import again if something is wrong?

Yes. The banner at the top of the board has Discard & re-import (company admins), which deletes every imported issue, comment, sprint and attachment and opens the Project step again — the key and name stay, but you can pick another board type (and environments), for example Delivery pipeline instead of Kanban. This works until the first change is made in the project — an issue, comment, move, sprint or doc, from the web or an AI agent. After that, importing is closed for good.