Code where you communicate.
Trigger cloud agents directly from Slack and return to a completed pull request.
swhitmore5m
dashboard is slow, 4s+ on first load. @cursor can you look into this?
CursorAPP3m
Found it — 3 sequential DB queries and no caching. Moved to parallel, added Redis, lazy-loaded the charts. Down to 380ms.
swhitmore2m
nice @eric can you review?
eric1m
lgtm @cursor merge it
CursorAPPnow
Merged to main.
Used by engineering teams at
@Cursor reads the thread, picks the work, ships a PR.
The integration handles routing and context so you only have to write the prompt.
Smart repo and model routing
Cursor reads your message, picks the right repository, and starts an agent on the right model. Name the repo or model in your prompt to override the defaults.
Thread context, automatically
Mention @Cursor in an existing thread and the agent reads the whole conversation as context. Bug reports, designs, error traces; no re-pasting.
Status updates and PR handoff
Agents react to your message while they run, then post back with a link to open the PR in GitHub or jump into Cursor desktop.
Set it up once for the whole team.
Configure defaults per channel, route keywords to the right repos, and keep privacy controls in your hands.
Channel settings
Set a default repository and base branch for any public channel so teammates can mention @Cursor without specifying either.
Routing rules
Map keywords to repositories or multi-repo environments so an agent boots with the right context every time.
Privacy mode
No training on your data by Cursor or LLM providers. Control whether agent summaries show in Slack Connect channels with external members.
Engineers ship from where the conversation happens.
A lot of our work comes from ideas and discussions in Slack. You can see the message, kick off @cursor in the same context, and you get a PR a few minutes later. This helps me avoid jumping between tools and context while the agent does the work.
Luke Bjerring Principal Engineer, Faire
Automations have made the repetitive aspects of my work easy to offload. By making automations to round up tasks, deal with doc updates, and respond to Slack messages, I can focus on the things that matter.
Tim Fall Senior Staff Software Engineer, Rippling
Questions & Answers
Go to the Cursor dashboard and click Connect next to Slack. After installing in Slack, you'll return to Cursor to connect GitHub, pick a default repository, and confirm privacy settings. Full walkthrough is in the Slack integration docs.
Cursor needs permissions to read mentions, channel and thread history for context, and to post status updates, completion notifications, and reactions. Each permission is listed with a rationale in the Slack integration docs.
Yes. Cursor picks the repository in this order: message content, recent agent activity, routing rules, channel default, and your personal default. Mention the repo in your message (for example, "@Cursor in cursor-app, fix the login bug") to force a specific one.
Yes. Privacy Mode is supported: no training on your data by Cursor or LLM providers. You can also control whether agent summaries appear in Slack Connect channels with external members. Privacy Mode (Legacy) is not supported because cloud agents need temporary code storage while running.
Mention @Cursor in the same thread to add follow-up instructions to the agent you own. To start a separate agent in a thread that already has one, use "@Cursor agent [prompt]". For multi-agent threads, use the context menu on the specific agent's message and pick Add follow-up.