When calling @cursor from slack, how to I instruct it to use my self hosted agent instead of cloud? I know I can do it from the web dashboard, but via slack it always spins a cloud vm instaed of using my personal environment.
Hey, thanks for the question.
Right now, the default Slack integration spins up a cloud VM. Support for self-hosted workers via Slack is something the team is working on, but it’s not fully implemented yet.
What you can try now:
- Make sure self-hosted agents are enabled in the Cloud Agents dashboard https://cursor.com/dashboard/cloud-agents#self-hosted-agents and that your worker is online.
- Try adding
@cursorin Slack withself_hosted=trueorprivate_worker=truein your message. There’s experimental support for this, but it can be unstable. - If you can, enable Use self-hosted workers globally in the Cloud Agents dashboard settings for all requests.
The team is aware of this limitation, and this is actively being worked on. There’s no exact ETA yet, but your request helps with prioritization.
Let me know if any of this works.
@deanrie hey, I got the same issue; It doesn’t work…
Hey, from the screenshots it looks like the pool is enabled and the worker is connected, so the issue is with the flag itself.
Right now the Slack integration only recognizes private_worker=true. The self_hosted=true syntax isn’t supported yet (they plan to add it as an alias, but it’s not released, and I can’t share a timeline).
Try one of these:
- In your message to
@Cursor, useprivate_worker=trueinstead ofself_hosted=true. - Or enable it globally in the Cloud Agents dashboard: https://cursor.com/dashboard/cloud-agents#self-hosted-agents. Then all Slack triggers will route to the self-hosted worker without needing a flag in every message.
Let me know which option works.