Jobs and runs

A job is a single execution of a script. A run is a single execution of a workflow — it groups together the jobs for each of the workflow’s nodes. Every time a workflow is triggered — manually, on a schedule, or via webhook — Orvanta creates a new run with its own isolated context, logs, and audit trail.

Run lifecycle

Queued → Running → Succeeded
                 → Failed
                 → Cancelled
                 → Awaiting approval  (when an approval step is active)

Status	Description
queued	The run has been accepted and is waiting for a worker.
running	At least one node is currently executing.
awaiting_approval	An approval step is waiting for human sign-off.
succeeded	All nodes completed without error.
failed	At least one node exited with a non-zero status or threw an exception.
cancelled	The run was manually cancelled before completion.

Run context

Each run carries a context — a key-value store that flows through the workflow. Nodes can read from and write to the context, allowing data to pass between steps without external state.

Context values are typed (string, number, boolean, JSON object) and are stored in the run record for full auditability.

Inspecting runs

Via the UI

Open any workflow in the dashboard and click the Runs tab. Select a run to see its node-by-node execution timeline, logs, and context snapshot.

Via the CLI

# List recent jobs
orvanta job list

# Get a job's details (for flows, the step tree)
orvanta job get <id>

# Fetch a job's logs
orvanta job logs <id>

Retention

Job logs and results are retained according to your instance’s configured retention policy. Self-hosted deployments can configure retention in the platform settings.

Related

• Workflows — How workflows are structured and triggered.
• Approval steps — Human approval step that puts a run in awaiting_approval.
• CLI: /job — Full CLI reference for jobs.