Detect blocked or silently stalled Linear issues

An Orkeste agent watches Linear for issues that look on track but have stopped moving: no commits, dependency unresolved, owner overloaded. It flags them for the engineering lead before they derail the sprint.

How the workflow runs

1. Monitor activity: Watches state changes, commits and comments per Linear issue.
2. Detect stall patterns: Identifies dependency stalls, owner overload, missing context.
3. Diagnose likely cause: Reads recent context and proposes a working hypothesis.
4. Eng lead approves alert: Lead reviews the flag and decides whether to escalate or coach.
5. Post to Slack DM: Approved alert is sent privately to the owner and the lead.

Human-in-the-loop: Alerts to owners are sent only after an engineering lead has approved the framing and recipient.

Works with your stack

• linear
• slack
• jira

Related workflows

• Summarize sprint progress in Linear automatically
• Triage incoming bugs in Jira automatically
• Alert on stalled deals in Salesforce

Questions about this workflow

How do you avoid alarming owners unnecessarily?

Alerts route privately to the lead first, who chooses to coach silently or send the alert to the owner.

Can it detect dependencies across teams?

Yes. The monitor follows cross-team blocked-by links and flags inter-team stalls explicitly.

What if an owner is on PTO?

PTO status is respected, alerts are deferred or rerouted to a designated backup.