Hey, thanks for the report. I get the frustration - cloud agents through Linear should work, not just burn money. Let’s check quickly and gather some data.
Similar issues on the forum:
- Agents hang on “Environment ready” → “Planning next moves” and after 30-50 minutes show “Ready for follow-up” without messages - server-side issue, team fixed it in November: Background Agent stuck on loading with small prompt
- Linear creates 2 agents instead of one - that’s a feature (Cursor runs two in parallel and picks the best), but yeah, it doubles the spend: Cursor Linear integration starting two background agents at once
Could you please check and share:
-
Integrations (Cursor - The AI Code Editor):
- Is GitHub connected and has access to the repo?
- Is Linear integration active?
- Is Default repository set in Settings > Features > Background Agents?
-
What’s happening:
- Screenshot from Linear issue - what status does Cursor show, any messages in the thread?
- Screenshot from Agents | Cursor - The AI Code Editor - can you see the agents, what’s their status (Starting / Planning / Ready)?
- Request ID of the hung agent (in Agents | Cursor - The AI Code Editor click on the agent → menu → Copy Request ID)
-
Configuration:
- Settings > Features > Background Agents > Personal Configuration - are you using Default Environment or a custom Dockerfile? (One user fixed it by removing a buggy dockerfile)
- Cursor version: Help > About Cursor > Copy
-
Temporary workaround:
- Try running the agent directly via Agents | Cursor - The AI Code Editor instead of Linear - if it works there, the issue is in the integration
- If it hangs there too - open Help > Toggle Developer Tools > Console, look for errors with “agent” or “linear”
Share the details - I’ll pass it to the team or we’ll find a fix.