Connecting data sources
Every widget, timesheet hour, and board row downstream of a source depends on it being connected correctly — this page is the map. It explains the two ways a connection attaches to your data, then links to each provider's specifics.
The Add-source wizard, step 1: choose a provider.
Two ways to connect
Almost every source is really two things: a shared connection (an Atlassian site, a GitHub/GitLab/Azure DevOps account, or a Microsoft Graph identity) and one or more project/repo syncs attached to it. The connection is created once and reused; each board then attaches its own subset of what that connection can see.
- A shared connection — Jira instances, GitHub/GitLab/Azure DevOps accounts, and the Microsoft Graph org-directory link are all reusable. Create one, and any board can attach a project or repo from it without re-entering credentials.
- A source attached to one board — the actual thing a board syncs (a Jira project key, a GitHub repo, a GitLab project, an ADO project) plus its board-specific settings: which group it lands in, which issue types are allowed, and its status mapping. The calendar source is the one connection that's board-scoped from the start — it isn't shared across boards at all.
Jira is read-only end to end. Deckgauge never writes an issue, a comment, or a status transition back to Jira — every call it makes is a read.
The shared lifecycle
- Create the connection — from a board's Sources tab (Add source) or the global
/sourcespage. Enter credentials and the connection is tested before it's saved. - Discover and pick — the wizard lists what the connection can see (projects, repos) so you choose rather than type blind.
- Attach and configure — pick a target group, which item types feed the board, and the status mapping.
- Sync — automatic on a schedule, or triggered manually. See When your data refreshes.
- Watch its health — a token going bad shows up as a badge and a reconnect banner, not a silent failure. See When a connection breaks.
Choose your source
| Source | What it's for |
|---|---|
| Jira | Issues, transitions, and sprint fields from a read-only Atlassian connection |
| GitHub | PRs, reviews, commits, issues, and milestones, with opt-in code intelligence per repo |
| GitLab | Merge requests, issues, reviews, and a separate commit-sync toggle |
| Azure DevOps | Work items always sync; PR/commit intelligence is strict opt-in |
| Microsoft Graph | Builds an org tree's reporting hierarchy from a root person down |
| Calendar | Pulls interview events into a recruitment board |
| Status mapping | How source-side statuses become board statuses (reference) |
| Sync & scheduling | Automatic vs manual sync, and retiring a decommissioned project |
| Connection health | Reading the health badge and reconnecting a broken source |
Related
- Boards — where synced and manual rows live side by side.
- Org trees — the hierarchy Microsoft Graph populates.
- Engineering intelligence — what code and issue data powers once it's connected.
Last updated