Tines¶
Tines stories report runs with one HTTP Request action at the end of the story. No SDK, no custom action: the standard HTTP Request action is the whole integration.
The action¶
Add an HTTP Request action as the story's final step:
| Setting | Value |
|---|---|
| URL | https://your-lumatrack-host/api/v1/runs |
| Method | POST |
| Content type | JSON |
| Headers | Authorization: Bearer <API key> (store the key as a Tines credential, never inline) |
Payload:
{
"automation": "phishing-triage",
"status": "success",
"source": "tines",
"external_id": "<<META.story_run_guid>>",
"units": "<<alerts.count>>"
}
external_id from the story run GUID makes retried runs idempotent.
units is optional: set it when one story run processes a variable batch
(alerts triaged, mailboxes swept) and value the automation
per unit.
Wire the failure path¶
A story that only reports success produces a number nobody should put in
front of a leadership review. Duplicate the action on the error path (or
in a story-level error handler) with "status": "failure" and a
failure_reason naming the root cause; the reliability chart and the
failure Pareto build from it.
What to use as the baseline¶
SOC automations usually replace analyst minutes: set the automation's manual baseline to the analyst time one triage takes (decompose it into baseline steps at the analyst's loaded rate for a figure that survives review), and let oversight minutes carry the spot-check time your team still spends.