UI Guide

DAG Detail Page

Clicking any workflow name opens its detail page (/dags/<id>) — this is where you inspect a single pipeline in depth: its schedule, code, run history, and task graph.

Header#

Every tab on this page shares the same header showing the DAG's key metadata:

Name + on/off switch — pause/activate this specific DAG (same as the list-page toggle).
Trigger button — manually start a new run of this DAG right now.
Schedule — the cron expression plus a human-readable translation.
Latest Run / Next Run — timestamps, with a status icon on the latest one.
Owner, Tags, Latest Dag Version — metadata about the current file/version.

Triggering a run#

Click Trigger to open the trigger dialog, titled Trigger DAG: <dag name>. If the DAG accepts runtime parameters, they're shown here as a form; if not, you'll see a note that it will be triggered with its default configuration.

Click Trigger to confirm — the new run kicks off immediately and the Grid view below refreshes automatically to show it in progress:

Grid view updating right after a trigger

The nine tabs#

Below the header sits a row of tabs. Here's what each one is for:

Grid#

A compact heatmap: one row per task, one column per recent run, colored by outcome. It's the fastest way to see "has this task been failing lately?" across many runs at once.

Success
Failed
Running
Queued
Skipped
Up For Retry
Upstream Failed

Overview#

The DAG's static metadata in one place: DAG ID, active/paused status, owner, schedule, latest version, next run, start date, source file path, tags, and a rendered description (if the DAG defines one).

Runs#

A reverse-chronological list of every run of this DAG, each showing start/end time, duration, any run parameters (hover the "Params" pill), and how long ago it ran. Click a run to open its Run Detail Page.

Calendar#

A month-view calendar where each day with runs shows colored dots for their outcomes — useful for spotting patterns like "this DAG always fails on weekends."

Tasks#

The static list of tasks that make up this DAG — operator type, trigger rule, retry count, and the state/time of each task's most recent execution.

Graph#

A visual node graph of task dependencies — arrows show which tasks must finish before the next one starts. Good for understanding a pipeline's shape at a glance.

Code#

The actual Python source file defining this DAG, syntax-highlighted (switchable between a dark and light theme). This is pulled live from the connected Git repository.

Diff#

Compares the currently-deployed code against a previous version, with additions/removals highlighted — useful for answering "what changed since this DAG last worked?"

Access#

Per-DAG permission overrides — grant a specific role extra access to this one DAG beyond what their role normally allows (e.g. give the Viewer role trigger rights on just one pipeline). The tab itself is visible to everyone; only Admins can add or remove entries here.