Is Grok Bot down right now?
The panel at the top of this page reads status.x.ai, where Cursor's status page moved on October 26 and which lists Grok Bot as its own component, and refreshes every minute while the page is open. If the Grok Bot component shows an incident, wait rather than reset anything. If it shows operational and only your account is affected, the likely cause is your account's shared computer or a Bot waiting on you, and the official recovery steps apply.
A green panel only means no incident has been declared. It says nothing about your account, and a declared incident can trail the first forum reports by hours.
Cursor's October 19 notice said that from October 26, 2026 its status updates would be published on status.x.ai, that status.cursor.com would stop sending notifications, and that incidents from the last 90 days could be found on status.x.ai. When we checked on October 27 the move was complete: every status.cursor.com address, including old incident links and the notice itself, redirected to status.x.ai, which lists Grok Bot and cursor.com as components and shows Cursor's incident history back to September. The incident links in this guide now open the status.x.ai history page.
Grok Bot signs in with your Cursor account, and the docs say the computers Bots work on run in Cursor's cloud. That is why the two components worth watching are Grok Bot, for the Bots and their computer, and cursor.com, for sign-in.
Documentation: SpaceXSI status page (Cursor's status page since October 26; Grok Bot and cursor.com components) ↗ Cursor status notice: status updates move to status.x.ai from October 26 (posted October 19) ↗ Grok Bot overview ↗ Official troubleshooting ↗
Where to check, in order
Declared incidents and user reports do not always arrive together. On September 16 the first post in the main unresponsive-Bots forum thread went up at 11:06 UTC and a Cursor staff member replied at 11:23 that they were investigating; the Grok Bot incident on the status page opened at 14:27 UTC. On a bad day, check the forum as well as the status page.
Documentation: SpaceXSI status page (Cursor's status page since October 26; Grok Bot and cursor.com components) ↗ Status history on status.x.ai (Cursor's incidents back to September) ↗ Status RSS feed (status.x.ai/feed.xml; Grok Bot items are prefixed [Grok Bot]) ↗ Cursor status notice: status updates move to status.x.ai from October 26 (posted October 19) ↗ Cursor forum: grok-bot tag ↗ Forum: some Bots became unresponsive (September 16, staff replies) ↗ Cursor status: Grok Bot degradation, September 16 ↗
| Source | What it covers | When to use it |
|---|---|---|
| Status page (status.x.ai) | Declared incidents for Grok Bot, cursor.com sign-in, Cloud Agents, Automations, the IDE and the CLI, alongside Grok on web, iOS, Android and X, the API and the console; Cursor's status page moved here on October 26 | First, every time |
| Status history (status.x.ai/history) | Past incidents with start and resolve times, including Cursor's incidents from before the move | To match a bad evening against a past incident |
| Status RSS feed (status.x.ai/feed.xml) | Every declared incident, including other products; Grok Bot items are prefixed [Grok Bot] | To be notified; Slack users can add it with /feed subscribe |
| Cursor forum, grok-bot tag | User reports and Cursor staff replies | When the status page is green but many people report the same symptom in the same hour |
Latest Grok Bot incidents (updated automatically)
This table is read from the incident history on status.x.ai, which now also holds Cursor's older status history, and refreshed every few minutes. It lists every incident from the last 30 days that named Grok Bot or cursor.com sign-in as affected, newest first, with times in UTC. The summaries below put the bigger ones in context.
| Opened (UTC) | Incident | Duration | Impact |
|---|---|---|---|
| Nov 6, 13:57 | Grok Bot is currently degraded ↗ · Grok Bot | 1 h 17 min | Minor |
| Nov 6, 11:49 | Computer use in Grok Bot fails for users in a few regions. ↗ · Grok Bot | 14 min | Minor |
| Nov 5, 00:34 | Degraded Performance - Grok Bot ↗ · Grok Bot | 1 h 15 min | Minor |
| Nov 2, 09:09 | Issues with source control integrations ↗ · IDE, Grok Bot | 1 h 15 min | Minor |
| Nov 1, 05:38 | Service Degradation - Billing and Usage Dashboard ↗ · cursor.com, Enterprise Dashboard | 5 h 15 min | Minor |
| Oct 31, 06:04 | Multiple services degraded ↗ · Automations, Cloud Agents, cursor.com, Grok Bot | 44 min | Minor |
| Oct 29, 23:44 | Auto Review Degradation ↗ · Grok Bot | 1 h 36 min | Minor |
| Oct 29, 09:01 | Degraded performance — Origin and Cloud Agents ↗ · Automations, Bugbot, Cloud Agents, Grok Bot, Origin | 17 min | Minor |
| Oct 28, 21:15 | Multiple services degraded ↗ · Automations, Bugbot, Cloud Agents, Grok Bot, Origin | 1 h 58 min | Major |
| Oct 27, 20:33 | Delayed responses in Grok Bot ↗ · Grok Bot | 10 min | Major |
| Oct 27, 00:19 | Payment System Sync Issues ↗ · cursor.com | 5 min | Minor |
| Oct 23, 23:39 | Investigating service outage - Grok models unavailable ↗ · Automations, Cloud Agents, CLI, cursor.com, IDE, Cursor IDE (iOS), Grok Bot | 41 min | Major |
| Oct 21, 15:46 | Investigating service degradation ↗ · cursor.com, Grok Bot | 1 h 54 min | Minor |
| Oct 19, 18:09 | Investigating service degradation — Grok Bot ↗ · Grok Bot | 19 min | Minor |
| Oct 19, 16:15 | Investigating service degradation — Grok Bot ↗ · Grok Bot | 1 h 7 min | Minor |
| Oct 14, 15:46 | Investigating service degradation — Grok Bot ↗ · Grok Bot | 5 h 18 min | Minor |
| Oct 13, 21:13 | Grok Bot routines not triggering ↗ · Grok Bot | 1 min | Major |
| Oct 13, 00:22 | Degraded performance and errors across Cursor services ↗ · Automations, Cloud Agents, CLI, cursor.com, IDE, Cursor IDE (iOS), Grok Bot, Origin, Bugbot | 2 h 46 min | Minor |
| Oct 9, 22:37 | Investigating service degradation — Cloud Agents and Grok Bot ↗ · Cloud Agents, Grok Bot | 31 min | Minor |
Documentation: Status history on status.x.ai (Cursor's incidents back to September) ↗ SpaceXSI status page (Cursor's status page since October 26; Grok Bot and cursor.com components) ↗
What happened in October 2026
October has been the busiest month so far. On October 3 Bots stopped answering in the early hours US Eastern time; a Cursor staff member wrote that the fix landed at about 3:10 AM ET, that messages sent while it lasted were not retried automatically, and that routines due between about 2:00 and 3:10 AM ET that day did not run. On October 6 a degradation affecting Cloud Agents and Grok Bot opened at 19:27 UTC, Grok Bot was marked restored at 21:25 UTC, and a second Cloud Agents incident listing Grok Bot followed. On October 7 Grok Bot's computers in certain regions were degraded for about an hour and a half.
October 8 brought several problems at once. Early in the day (UTC) every Bot on some accounts answered Bot failed to respond; staff said it was a temporary issue on their side and resolved by about 09:55 UTC. From about 13:30 UTC some Bots could not reach their model because of a capacity limit, which staff said cleared at about 21:00 UTC; the status page listed Grok Bot in two incidents about errors and rate limiting on a model used across Cursor. Replies also failed for some users in one region, including Hong Kong, until a fix that evening. Scheduled routines were held back, ran late or were marked Failed from about 10 AM ET on October 8 into October 9, with no status entry; staff said the resolved notice on the status page that evening referred to a different issue, and that missed runs would not be replayed. On October 9 from 22:37 to 23:09 UTC some new Cloud Agents and Grok Bot sessions failed to start.
Not every problem gets a status entry. Cursor staff confirmed in the forum that since October 1 some Grok Bot computers can only make outbound HTTPS connections, so mail, database, SSH and FTP connections time out; they called it unintended and said resetting does not help. More accounts reported it after service-side computer refreshes on October 7 and 8, new reports were still arriving on October 14, and Cursor had not posted a fix in the thread by then.
A wider incident followed in the early hours of October 13 (UTC): the status page reported errors or failed requests across the IDE, Cloud Agents, Automations, Grok Bot and Origin starting at 00:22. Cursor said a fix was implemented by 00:53, that Grok Bot specifically had recovered and was operating normally by 01:49, and marked the whole incident resolved at 03:07 once Cloud Agents, which lagged behind the other services, caught up.
Routines failed twice more later on October 13. A Cursor staff member wrote in the forum that scheduled routines were held back again for a few hours that morning (US time) and had been going through since about 16:00 UTC, and that missed runs would not be replayed. Separately, Cursor posted a status notice after the fact: between 10:52 AM and 1:29 PM PT (17:52 to 20:29 UTC), many routines started by webhooks, Slack, git, email and Linear events did not run. Webhook callers may have seen an HTTP 400 error reading Automation run rate limit exceeded, while other triggers failed without an error. Cursor said users' own usage did not cause it and that events sent in that window were not processed and will not be retried, so resend anything important. That notice was published already resolved and lists no affected component.
The second half of the month has been quieter but not clean. On October 14 from 15:46 to 21:04 UTC Cursor reported a Grok Bot degradation that stopped some users reaching their computers, and said at 20:08 that it had identified the cause. On October 19 Grok Bot routines were delayed for about an hour (16:15 to 17:21 UTC), and shortly afterwards Grok Bot and grok.com showed errors about billing, plans and usage for 19 minutes. On October 21 errors on billing dashboards, usage and plan changes listed both cursor.com and Grok Bot for almost two hours, with a fix in place after about ten minutes. The largest was a major outage from 23:39 UTC on October 23 to 00:20 UTC on October 24, when Grok models were unavailable across Grok Bot, cursor.com, the IDE, the CLI, Cloud Agents and Automations; Cursor said a fix was implemented at 23:50. The first Grok Bot incident on the new status.x.ai page followed on October 27: from 20:33 to 20:43 UTC some Grok Bot responses were delayed, marked major with only Grok Bot listed, and Cursor said responses were back to normal when it closed the incident ten minutes later. A larger one followed on October 28: from 21:15 to 23:13 UTC Cursor reported failures across Cloud Agents, Grok Bot and Origin, then Automations and BugBot as well, marked major; it said a fix was in place at 22:06 and services had recovered by 22:40, and closed the incident about half an hour after that. October 29 brought two shorter, minor ones: from 09:01 to 09:18 UTC degraded performance in Cloud Agents and Origin, with Grok Bot among the listed components, where users could see slow responses, timeouts or retried requests; and from 23:44 UTC to 01:20 UTC on October 30 a degradation in Auto Review, with only Grok Bot listed, so approvals that Auto Review normally checks may have stalled or behaved oddly in that window. Another minor one followed on October 31: from 06:04 to 06:47 UTC Cursor reported multiple services degraded, listing Grok Bot, Cloud Agents, Automations and cursor.com, and closed it as resolved without posting a cause.
Grok Bot 0.71.2 (October 14) changed how some of these moments look in the app: while the Bot's computer is asleep it now says Your computer is asleep instead of Couldn't refresh or a connection error, a working Bot reconnects to the new computer when its computer updates, and a computer that fails to wake is reported as one that couldn't start rather than one still starting. If you are on an older version, update before deciding an error is an outage.
Documentation: Status: delayed responses in Grok Bot, October 27 ↗ Status: multiple services degraded, including Grok Bot, October 28 ↗ Status: degraded performance in Cloud Agents and Origin, listing Grok Bot, October 29 ↗ Status: Auto Review degradation in Grok Bot, October 29 to 30 ↗ Status: multiple services degraded, including Grok Bot, October 31 ↗ Status: billing and usage dashboard degradation, cursor.com and Enterprise Dashboard, November 1 ↗ Status: issues with source control integrations, IDE and Grok Bot, November 2 ↗ Status: degraded performance in Grok Bot, November 5 ↗ Status: computer use in Grok Bot fails in a few regions, November 6 ↗ Status: Grok Bot degraded, November 6 ↗ Forum: Grok Bot weekly usage meters jump on November 1 (user reports) ↗ Cursor status: Grok Bot and Review Agents degradation, October 3 ↗ Forum: Grok Bots failing to respond again (staff reply on the October 3 fix) ↗ Cursor status: Cloud Agents and Grok Bot degradation, October 6 ↗ Cursor status: Cloud Agents degradation listing Grok Bot, October 6 (late) ↗ Forum: Bot failed to respond on every Bot (staff replies, October 6 to 9) ↗ Cursor status: computers in certain regions degraded, October 7 ↗ Cursor status: model rate limiting listing Grok Bot, October 8 ↗ Cursor status: model errors listing Grok Bot, October 8 (evening UTC) ↗ Forum: replies failing for some users in one region (staff fix, October 9) ↗ Forum: all Grok Bot routines failed to run (staff replies, October 8) ↗ Cursor status: new Cloud Agents and Grok Bot sessions failing to start, October 9 ↗ Forum: outbound non-HTTPS traffic blocked since October 1 (staff reply) ↗ Cursor status: widespread degradation across the IDE, Cloud Agents, Automations, Grok Bot and Origin, October 13 ↗ Cursor status: Grok Bot routines not triggering, October 13 ↗ Cursor status: some users unable to reach their computers, October 14 ↗ Cursor status: Grok Bot routines delayed, October 19 ↗ Cursor status: Grok Bot billing, plans and usage errors, October 19 ↗ Cursor status: billing, usage and plan-change errors, October 21 ↗ Cursor status: outage, Grok models unavailable, October 23 to 24 ↗ Grok Bot changelog 0.71.2 (October 14): asleep and safety-check messages, wake in place ↗
| Date (UTC) | What the status page said | Duration |
|---|---|---|
| October 3, 06:50 to 07:26 | Degradation affecting Grok Bot and Review Agents | About 35 minutes |
| October 6, 19:27 to 22:06 | Degradation affecting some Cloud Agents and Grok Bot features; Grok Bot marked restored at 21:25 | About 2 hours for Grok Bot |
| October 6, 22:48 to 23:04 | Cloud Agents degradation, with Grok Bot listed as affected | 16 minutes |
| October 7, 15:59 to 17:25 | Degradation affecting Grok Bot's computers in certain regions | About 1.5 hours |
| October 8, 13:54 to 15:31 | Rate limiting (429 errors) on a model used across Cursor, with Grok Bot listed | About 1.5 hours |
| October 8, 19:52 to 20:51 | Elevated errors on the same model, with Grok Bot listed | About 1 hour |
| October 9, 22:37 to 23:09 | Some new Cloud Agents and Grok Bot sessions may fail to start | About 30 minutes |
| October 13, 00:22 to 03:07 | Widespread errors across the IDE, Cloud Agents, Automations, Grok Bot and Origin; Grok Bot itself recovered by 01:49 | About 2.75 hours overall, about 1.5 hours for Grok Bot |
| October 13, 17:52 to 20:29 | Many Grok Bot routines started by webhooks, Slack, git, email and Linear events did not run; posted after it ended, with no component listed | About 2.5 hours |
| October 14, 15:46 to 21:04 | Grok Bot degradation preventing some users from reaching their computers; cause identified at 20:08 | About 5.25 hours |
| October 19, 16:15 to 17:21 | Grok Bot degradation where routines may be delayed | About 1 hour |
| October 19, 18:09 to 18:28 | Grok Bot and grok.com errors related to billing, plans and usage | 19 minutes |
| October 21, 15:46 to 17:40 | Billing dashboard errors on billing, usage and plan changes, with cursor.com and Grok Bot listed; fix in place at 15:56 | About 2 hours |
| October 23 to 24, 23:39 to 00:20 | Outage: Grok models unavailable, with Grok Bot, cursor.com, the IDE, the CLI, Cloud Agents and Automations listed (marked major) | About 40 minutes |
| October 27, 20:33 to 20:43 | Delayed responses in Grok Bot, with only Grok Bot listed (marked major); the first incident posted on status.x.ai after the move | 10 minutes |
| October 28, 21:15 to 23:13 | Multiple services degraded: failures across Cloud Agents, Grok Bot and Origin, then also Automations and BugBot (marked major); a fix was in place at 22:06 and services had recovered by 22:40 | About 2 hours |
| October 29, 09:01 to 09:18 | Degraded performance in Cloud Agents and Origin, with Grok Bot, Automations, BugBot and Review Agents listed: slow responses, timeouts or retried requests | 17 minutes |
| October 29 to 30, 23:44 to 01:20 | Auto Review degradation, with only Grok Bot listed | About 1.5 hours |
| October 31, 06:04 to 06:47 | Multiple services degraded, with Grok Bot, Cloud Agents, Automations and cursor.com listed (marked minor) | About 45 minutes |
| November 1, 05:38 to 10:53 | Degradation affecting the billing and usage dashboard, with cursor.com and the Enterprise Dashboard listed but not Grok Bot (marked minor); forum users tied it to Grok Bot weekly usage meters that jumped the same morning, and many meters stayed high after it was marked resolved | About 5.25 hours |
| November 2, 09:09 to 10:24 | Issues with source control provider integrations in Cursor and Grok Bot, with the IDE and Grok Bot listed (marked minor) | About 1.25 hours |
| November 5, 00:34 to 01:49 | Degraded performance: errors with Grok Bot, with only Grok Bot listed (marked minor); the update says only that the issue was identified and resolved | About 1.25 hours |
| November 6, 11:49 to 12:03 | Computer use in Grok Bot failing for users in some regions, with other Grok Bot features working normally and only Grok Bot listed (marked minor) | About 15 minutes |
| November 6, 13:57 to 15:15 | Grok Bot degraded, with only Grok Bot listed (marked minor); the updates say only that the team was investigating and that it was resolved | About 1.25 hours |
Grok Bot incidents in September 2026
These are the incidents on Cursor's status page (now redirected to status.x.ai) in September that listed Grok Bot as affected, plus one cursor.com sign-in incident. Times are UTC, from when Cursor opened each incident to when it marked it resolved. The history also shows Grok Bot incidents on August 18, when connectors hung, and August 27.
In the forum, a Cursor staff member wrote on September 17 that services should be back after about 24 hours of disruption, although two more Grok Bot incidents followed on September 18. Staff closed the main thread on September 21, saying things were stable, and asked people to open a new thread for fresh reports. Five shorter incidents followed between September 25 and 29. Our not-responding guide describes what those days looked like from the user side.
Documentation: Status history on status.x.ai (Cursor's incidents back to September) ↗ Cursor status: degradation across Grok models and Grok Bot, September 3 ↗ Cursor status: routine trigger fires dropped, September 3 to 4 ↗ Cursor status: Grok Bot computers unreachable, September 7 ↗ Cursor status: service degradation including Grok Bot, September 15 ↗ Cursor status: Grok Bot degradation, September 16 ↗ Cursor status: Grok Bot degradation, September 17 ↗ Cursor status: Cloud Agents and Automations degradation, September 17 ↗ Cursor status: Grok Bot degradation, September 18 (morning UTC) ↗ Cursor status: Grok Bot degradation, September 18 (evening UTC) ↗ Cursor status: sign-in and dashboard errors, September 18 ↗ Cursor status: computers could not be created, reached or reset, September 25 ↗ Cursor status: Grok Bot degradation, September 25 to 26 ↗ Cursor status: Grok Bot degradation, September 28 ↗ Cursor status: Cloud Agents and Grok Bot degradation, September 29 ↗ Cursor status: xAI model degradation, September 29 ↗ Forum: some Bots became unresponsive (September 16, staff replies) ↗
| Date (UTC) | What the status page said | Duration |
|---|---|---|
| September 3, 13:41 to 17:07 | Degradation affecting all Grok models, Automations, Cloud Agents, Grok Bot and Review Agents (marked critical) | About 3.5 hours |
| September 3 to 4, 23:42 to 00:42 | Degraded performance for Grok Bot routine triggers; some trigger fires may be dropped | 1 hour |
| September 7, 20:32 to 21:37 | Some users unable to reach their Grok Bot computer; Recover or Reset may fail; data not affected | About 1 hour |
| September 15, 21:26 to 22:00 | Degradation across Cursor components, including Grok Bot | About 35 minutes |
| September 16, 14:27 to 20:57 | Grok Bot degradation: degraded performance or bots that don't respond | About 6.5 hours |
| September 17, 12:30 to 14:34 | The same Grok Bot degradation | About 2 hours |
| September 17, 15:20 to 20:04 | Cloud Agents and Automations degradation, with Grok Bot listed as affected | About 4.75 hours |
| September 18, 11:37 to 14:33 | Grok Bot degradation: degraded performance or bots that don't respond | About 3 hours |
| September 18, 17:19 to 23:02 | Grok Bot service degradation | About 5.75 hours |
| September 18, 23:02 to 23:13 | Sign-in and dashboard errors on cursor.com; existing sessions kept working | 11 minutes |
| September 25, 05:17 to 06:51 | Grok Bot computers could temporarily not be created, reached, or reset | About 1.5 hours |
| September 25 to 26, 23:47 to 00:32 | Grok Bot service degradation | About 45 minutes |
| September 28, 20:23 to 20:36 | Grok Bot service degradation | 13 minutes |
| September 29, 14:25 to 18:05 | Degradation affecting Cloud Agents and Grok Bot | About 3.75 hours |
| September 29, 19:43 to 20:18 | Degradation affecting xAI models: errors, slow responses or rate limiting, with Grok Bot listed | About 35 minutes |
An outage, or just your account?
Every Bot on an account shares one cloud computer, so a problem with that computer looks like an outage: all your Bots stop together. The difference is whether other people are affected at the same time. Use the status page and the forum to answer that before you change anything.
Documentation: Official troubleshooting ↗ Grok Bot FAQ ↗ SpaceXSI status page (Cursor's status page since October 26; Grok Bot and cursor.com components) ↗ Grok Bot changelog (Wake button in 0.69.0 to 0.70.2, October 11) ↗ Grok Bot changelog 0.71.2 (October 14): asleep and safety-check messages, wake in place ↗ Forum: all my Grok Bots disappeared after the X Subscription switch (staff replies, October 21) ↗
| What you see | Most likely | What to do |
|---|---|---|
| Several Bots stop at once and the status page shows a Grok Bot incident | A service outage | Wait; do not Reset. Note the time and your app version |
| Every Bot stops but the status page is green and the forum is quiet | Your account's shared computer is stuck or unreachable | Retry, restart the app, Recover computer, then Update Bot's Computer in Settings → Computer; Reset only last |
| The app says Your computer is asleep (older versions showed Couldn't refresh or a connection error) | An idle computer, not an outage; 0.71.2 replaced the misleading messages with this one | On 0.71.2 or later, click the Computer is asleep preview to wake it in place (it replaced the separate Wake button); switching chats does not wake it |
| "Blocked by a safety check" with "Try a different message" | A safety check blocked that Bot's reply, not an outage; 0.71.2 shows this instead of a generic error | Rephrase, or start the work with another Bot; see the not-responding guide |
| One Bot is silent while others reply | That Bot is waiting on a question, approval, login, CAPTCHA or secret request | Open the computer view, answer it, or send a short redirect or "Stop now" |
| Bots stop after a busy week, or show a usage message | Weekly usage is used up or the on-demand spending limit is reached | Check Usage & Billing; see our weekly usage limit guide |
| Sign-in never completes | A cursor.com sign-in incident, an SSO step, or Legacy Privacy Mode | Check the cursor.com component, then the sign-in checks |
| A scheduled routine did not run | A routine-trigger incident, or a disabled or misconfigured routine | Compare the time with the status history, then run the routine checklist |
| Bots are missing, or the app says your plan has no Grok Bot, after switching plans | An account split during the switch to The X Subscription, not an outage (Cursor staff, October 21) | Do not reset, delete or buy a second plan; follow the account-switch steps |
| Grok on grok.com or in X is failing too | A wider xAI problem | Check status.x.ai as well |
What to do during an outage
The goal is to avoid making things worse. User-side recovery rarely helps during a declared incident, and the one destructive option, Reset, can discard recent work.
Documentation: Grok Bot FAQ ↗ Official troubleshooting ↗ Cursor status: Grok Bot computers unreachable, September 7 ↗ Cursor status: routine trigger fires dropped, September 3 to 4 ↗ Status RSS feed (status.x.ai/feed.xml; Grok Bot items are prefixed [Grok Bot]) ↗ Cursor status notice: status updates move to status.x.ai from October 26 (posted October 19) ↗ Forum: all Grok Bot routines failed to run (staff replies, October 8) ↗ Cursor status: Grok Bot routines not triggering, October 13 ↗ Cursor status: Grok Bot routines delayed, October 19 ↗
- Leave running work alone. Bot work runs on the cloud computer, and the FAQ says closing the app, laptop or phone does not stop a background turn or routine.
- Do not use Reset. It restores the last saved snapshot and can lose recent or unsynced work, and during the September 7 incident Cursor said Recover or Reset might fail anyway.
- Subscribe to status updates so you know when the incident is marked resolved: from October 26 Cursor publishes them on status.x.ai (RSS at status.x.ai/feed.xml), and status.cursor.com stops sending notifications.
- Once it is resolved, open each Bot, read its last message, and resend anything that did not go through; Cursor staff said messages sent during the October 3 incident were not retried automatically.
- Check routines that were due during the incident window and run missed ones by hand, for example by asking the Bot to run them; the September 3 to 4 incident dropped some routine trigger fires, routines due during the October 3 incident did not run, runs missed on October 8 and 13 were not replayed, webhook, Slack, git, email and Linear events sent during the October 13 routine-trigger incident must be sent again, and routines were delayed on October 19.
- If Bots still do not respond after the incident is resolved, treat it as your computer and follow the recovery order in our not-responding guide.
- Before contacting support, collect your Grok Bot version and operating system.
- Copy the exact error message and the Bot or routine name.
- Note the approximate time and your time zone, and any request or conversation ID shown.
- Record whether Retry, an app restart, or updating Grok Bot's computer changed anything.
- Leave out passwords, one-time codes, private keys and secret values.
Quick answers
Is there an official Grok Bot status page?
Yes, on status.x.ai. Cursor's October 19 notice said that from October 26, 2026 its status updates would be published there, and that has happened: status.x.ai now lists Grok Bot as its own component next to cursor.com, Cloud Agents, Automations, the IDE and the CLI, alongside the Grok apps, the API and the console, its history page carries Cursor's past incidents back to September, and the panel at the top of this page reads it live. When we checked on October 27, every status.cursor.com address, including old incident links, redirected to status.x.ai, so links in this guide now point there.
Why is the status page green when my Bots are not responding?
Two reasons. The status page reports incidents Cursor has declared, not the state of your account, and every Bot on an account shares one cloud computer, so a stuck computer stops them all at once without affecting anyone else. And declared incidents can lag or never appear: on September 16 the first forum report came about three hours before the Grok Bot incident was opened, and the October 8 routine failures had no status entry at all. Check the forum's grok-bot tag, then follow the official recovery order.
Do I need to reset my computer after an outage?
Usually not. After the October 3, 6 and 8 incidents, Cursor staff said the fix was on their side, that no reset or reinstall was needed, and that any message still showing Bot failed to respond should simply be sent again. Messages sent during the October 3 incident were not retried automatically, and routines missed on October 3, October 8 and October 13 were not replayed, so run those by hand; events sent to event-triggered routines during the October 13 incident were not processed, so resend them. Routines were also delayed for about an hour on October 19, so check any that were due then. Reset restores the last saved snapshot and can lose recent work.
Will I lose my Bots or files during an outage?
Not from the outage itself, going by what Cursor has published. The September 7 status notice said data was not affected, and the troubleshooting page says Bot profiles and saved conversations are not necessarily lost when the computer is temporarily unreachable. The risk is Reset: it restores the last saved snapshot and can lose recent or unsynced work, so leave it until last.
How do I get notified when Grok Bot is down?
Move your subscription to status.x.ai. Cursor's October 19 notice says that from October 26 status.cursor.com stops sending notifications. Email subscriptions move over automatically, with mail sent from status@email.api.x.ai; the RSS feed to use is status.x.ai/feed.xml, which Slack users can add with /feed subscribe; and SMS, Microsoft Teams and webhook notifications stop on October 26, so re-subscribe by email or RSS. The feeds carry every declared incident, including ones for other products, so filter for Grok Bot and cursor.com if you only care about sign-in and Bots.
Sources & next steps
Capabilities are grounded in the documentation below. The workflow design and acceptance checks are editorial suggestions.
- SpaceXSI status page (Cursor's status page since October 26; Grok Bot and cursor.com components) ↗
- Status history on status.x.ai (Cursor's incidents back to September) ↗
- Status RSS feed (status.x.ai/feed.xml; Grok Bot items are prefixed [Grok Bot]) ↗
- Status: billing and usage dashboard degradation, cursor.com and Enterprise Dashboard, November 1 ↗
- Status: Grok Bot degraded, November 6 ↗
- Status: computer use in Grok Bot fails in a few regions, November 6 ↗
- Status: degraded performance in Grok Bot, November 5 ↗
- Status: issues with source control integrations, IDE and Grok Bot, November 2 ↗
- Forum: Grok Bot weekly usage meters jump on November 1 (user reports) ↗
- Status: multiple services degraded, including Grok Bot, October 31 ↗
- Status: Auto Review degradation in Grok Bot, October 29 to 30 ↗
- Status: degraded performance in Cloud Agents and Origin, listing Grok Bot, October 29 ↗
- Status: multiple services degraded, including Grok Bot, October 28 ↗
- Cursor status notice: status updates move to status.x.ai from October 26 (posted October 19) ↗
- Status: delayed responses in Grok Bot, October 27 ↗
- Cursor status: outage, Grok models unavailable, October 23 to 24 ↗
- Cursor status: billing, usage and plan-change errors, October 21 ↗
- Cursor status: Grok Bot billing, plans and usage errors, October 19 ↗
- Cursor status: Grok Bot routines delayed, October 19 ↗
- Cursor status: some users unable to reach their computers, October 14 ↗
- Cursor status: Grok Bot routines not triggering, October 13 ↗
- Cursor status: widespread degradation across the IDE, Cloud Agents, Automations, Grok Bot and Origin, October 13 ↗
- Cursor status: degradation across Grok models and Grok Bot, September 3 ↗
- Cursor status: routine trigger fires dropped, September 3 to 4 ↗
- Cursor status: Grok Bot computers unreachable, September 7 ↗
- Cursor status: service degradation including Grok Bot, September 15 ↗
- Cursor status: Grok Bot degradation, September 16 ↗
- Cursor status: Grok Bot degradation, September 17 ↗
- Cursor status: Cloud Agents and Automations degradation, September 17 ↗
- Cursor status: Grok Bot degradation, September 18 (morning UTC) ↗
- Cursor status: Grok Bot degradation, September 18 (evening UTC) ↗
- Cursor status: sign-in and dashboard errors, September 18 ↗
- Cursor status: computers could not be created, reached or reset, September 25 ↗
- Cursor status: Grok Bot degradation, September 25 to 26 ↗
- Cursor status: Grok Bot degradation, September 28 ↗
- Cursor status: Cloud Agents and Grok Bot degradation, September 29 ↗
- Cursor status: xAI model degradation, September 29 ↗
- Cursor status: Grok Bot and Review Agents degradation, October 3 ↗
- Cursor status: Cloud Agents and Grok Bot degradation, October 6 ↗
- Cursor status: Cloud Agents degradation listing Grok Bot, October 6 (late) ↗
- Cursor status: computers in certain regions degraded, October 7 ↗
- Cursor status: model rate limiting listing Grok Bot, October 8 ↗
- Cursor status: model errors listing Grok Bot, October 8 (evening UTC) ↗
- Cursor status: new Cloud Agents and Grok Bot sessions failing to start, October 9 ↗
- Forum: Grok Bots failing to respond again (staff reply on the October 3 fix) ↗
- Forum: Bot failed to respond on every Bot (staff replies, October 6 to 9) ↗
- Forum: replies failing for some users in one region (staff fix, October 9) ↗
- Forum: all Grok Bot routines failed to run (staff replies, October 8) ↗
- Forum: outbound non-HTTPS traffic blocked since October 1 (staff reply) ↗
- Cursor forum: grok-bot tag ↗
- Forum: some Bots became unresponsive (September 16, staff replies) ↗
- Official troubleshooting ↗
- Grok Bot FAQ ↗
- Grok Bot overview ↗
- Grok Bot changelog (Wake button in 0.69.0 to 0.70.2, October 11) ↗
- Grok Bot changelog 0.71.2 (October 14): asleep and safety-check messages, wake in place ↗
- Forum: all my Grok Bots disappeared after the X Subscription switch (staff replies, October 21) ↗