Skip to main content

Runs, Reports, And Exports

Runs​

A run is a single execution of a preset against one or more source documents.

Runs move through statuses such as pending, running, paused, cancelled, failed, completed, and completed-with-errors.

What A Run Stores​

The backend stores run metadata in the per-user encrypted DB and stores run artifacts on disk under:

data/user_{uuid}/runs/{run_id}/

That run directory can contain generated markdown files, reports, exports, and related artifacts.

Reports​

Users can request a run report at:

GET /api/runs/{run_id}/report

The backend reuses report.html if it already exists. If it does not exist yet, the backend builds it from normalized run-result tables.

Exports​

Users can generate and download a ZIP export:

  • POST /api/runs/{run_id}/export
  • GET /api/runs/{run_id}/export

The backend now reuses an existing export.zip instead of rebuilding it on every read.

Logs​

Run logs are stored separately from the main user DB in a per-user sidecar logs database.

Users can access structured run logs through:

  • GET /api/runs/{run_id}/logs
  • GET /api/runs/{run_id}/logs/count