Connecting Jira

A Jira connection is an Atlassian site plus an API token, reused across every board that pulls from it. Deckgauge only ever reads from Jira — there is no write path back.

deckgauge · Add source
JIJira
GHGitHub
GLGitLab
AZAzure DevOps

The Add-source wizard's provider step, Jira selected.

What it does, and where the data comes from

Deckgauge is a read-only Jira client. Creating a connection stores a site URL, an account email, and an API token; every call it makes afterward is a GET — testing the token, listing projects, pulling issues. It never opens, transitions, comments on, or edits a Jira issue.

FieldValue
Instance NameA label for the connection, e.g. My Jira Cloud
Jira Cloud URLhttps://your-domain.atlassian.net — must be the real Atlassian host, not a display domain (see below)
EmailThe account email the API token belongs to
API TokenCreated at id.atlassian.com/manage-profile/security/api-tokens; sent as HTTP Basic auth (email + token)
Project Keyse.g. PLAT, WEB — added one at a time, discovered live once the token is valid

How to connect a Jira instance

  1. Create an Atlassian API token at id.atlassian.com/manage-profile/security/api-tokens and copy it. Jira API tokens carry no scopes — Deckgauge sees exactly the projects your account can already see.
  2. Add source → Jira — enter the Instance Name, Jira Cloud URL, Email, and API Token, then Create & test. The test calls Jira's /rest/api/3/myself with your credentials.
  3. Pick the project key(s) from the live-discovered list (Jira's /rest/api/3/project), or type a key directly.
  4. Review & attach — the project is attached to the board's Sources tab and starts syncing.

What syncs

DataNotes
IssuesEvery issue under the project key(s) you chose, filtered to the issue types you allow on the board's Source zone
TransitionsFull status-change history per issue — the input to cycle-time, flow, and timesheet in-progress calculations
Sprint fieldsSprint name/state, auto-detected by shape rather than a hardcoded custom-field id (Jira's numeric field id for Sprint varies per site) — feeds Velocity and Iteration Planning Accuracy
Due dateThe issue's Jira duedate, written onto the board row's Due date field on every sync — until someone edits it by hand, which takes that field out of sync's control (see Manual edits)
WorklogsOptional, off by default — enable it on the connection if you want logged-time data

Under the hood

Jira Cloud sites answer on two hosts: the canonical *.atlassian.net API host, and — if your organization uses one — an SSO vanity display domain (something like jira.yourcompany.com). The vanity host renders the Jira UI fine, but it silently discards HTTP Basic credentials, so a perfectly valid API token gets a 401 against it.

The vanity-domain trapIf your connection test fails with a 401 and you're sure the token is right, you're almost certainly pointed at the vanity domain. Deckgauge detects this: on a 401/403 it makes one unauthenticated call to Jira's public serverInfo endpoint, compares hosts, and — when they differ — shows a banner with the real host and a Use this URL button that repoints the connection and retests automatically.

If it looks wrong

SymptomCauseFix
Test fails with 401, token looks correctURL points at an SSO vanity/display domain, not the real Atlassian hostClick Use this URL on the banner, or manually switch to *.atlassian.net
No Jira issues appear on the boardThe issue's type isn't in the board's allowed issue types for this sourceOpen the board's Source zone and add the type to the chip list
Sprint-based widgets are blankThe project doesn't use Jira sprints, or issues haven't synced since sprint data was addedConfirm the project runs sprints in Jira, then re-sync
Retired project keeps accruing timesheet hoursProject stopped syncing but its items never reached a terminal statusMark it in Retired projects with a cutoff date

Related

Last updated