# Cursor Live Task Allow List does not include Grok 4.7 **URL:** \ **Category:** Bug Reports\ **Tags:** subagents, grok\ **Created:** [September 21, 2026, 8:11pm UTC](https://forum.cursor.com/t/cursor-live-task-allow-list-does-not-include-grok-4-7/172552 "2026-09-21T20:11:09Z")\ **Posts on this page:** 7\ **Page:** 1
**Author:** ![Zachariah](https://avatars.discourse-cdn.com/v4/letter/z/c4cdca/32.png) [@Zachariah](https://forum.cursor.com/u/Zachariah)\ **Post date:** [September 21, 2026, 8:11pm UTC](https://forum.cursor.com/t/cursor-live-task-allow-list-does-not-include-grok-4-7/172552/1 "2026-09-21T20:11:09Z")
### Where does the bug appear (feature/product)? Cursor IDE ### Describe the Bug When running an agent as Grok 4.7 Extra High, child workers cannot be spawned as Grok 4.7 because they are not in the “Live Task Allow List” From the model’s own diagnostic of this problem: This chat’s picker is already Grok 4.7, reasoning xhigh, Fast off. The CLI id for that variant is grok-4.7-xhigh. The live Task allowlist is still: composer-2.5-fast cursor-grok-4.6-high-fast gemini-3.6-flash-high gpt-5.6-sol-low Probe Result Model omitted Auto Legacy slug grok-4.7-xhigh Auto Frontmatter grok-4.7-xhigh Composer Allowlisted cursor-grok-4.6-high-fast Cursor Grok 4.6 The exact parameter form grok-4.7[context=256k,reasoning\_effort=xhigh,fast=false] was rejected with the same four-slug list. A custom subagent type was rejected because Task only accepts the built-in types. Changing the saved enabled-model list from grok-4.6 to grok-4.7 did not refresh that list, so I restored the original list. ### Steps to Reproduce Try to run a subagent/worker from within an agent call using Grok 4.7, and have the caller/orchestrator agent check which model actually got called - defaults back to composer. ### Expected Behavior Grok 4.7 should call Grok 4.7 workers. ### Operating System Windows 10/11 ### Version Information Version: 3.21.16 (system setup) VS Code Extension API: 1.128.0 Commit: 8ae78e8eee1e63479c7e0504b664bc0a80c68000 Date: 2026-09-19T06:43:22.276Z Layout: IDE Build Type: Stable Release Track: Default Electron: 42.10.0 Chromium: 148.0.7778.280 Node.js: 24.18.1 V8: 14.8.178.38-electron.0 xterm.js: 6.1.0-beta.291 OS: Windows\_NT x64 10.0.19045 ### Does this stop you from using Cursor Yes - Cursor is unusable ---
**Author:** ![Zachariah](https://avatars.discourse-cdn.com/v4/letter/z/c4cdca/32.png) [@Zachariah](https://forum.cursor.com/u/Zachariah)\ **Post date:** [September 22, 2026, 3:17am UTC](https://forum.cursor.com/t/cursor-live-task-allow-list-does-not-include-grok-4-7/172552/6 "2026-09-22T03:17:51Z")
Currently using a custom MCP server to run Cursor CLI workers at any thinking level I want. It’s a hack, and I have to tell the agent to wait for the responses, otherwise it ends the turn. Would love for this to be resolved. ---
**Author:** ![mohitjain](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/mohitjain/32/100229_2.png) [@mohitjain](https://forum.cursor.com/u/mohitjain)\ **Post date:** [September 22, 2026, 6:57am UTC](https://forum.cursor.com/t/cursor-live-task-allow-list-does-not-include-grok-4-7/172552/8 "2026-09-22T06:57:03Z")
Hey @Zachariah, To get real Grok 4.7 workers today, delegate from a Cloud Agent - its subagents can be pinned to Grok 4.7 at any effort level (locally the Task list only exposes the Grok 4.6 entry, which is why editing your enabled models didn’t help). This isn’t anything on your end, and it’s an issue we’re tracking - I’ll post back here when there’s an update. ---
**Author:** ![Zachariah](https://avatars.discourse-cdn.com/v4/letter/z/c4cdca/32.png) [@Zachariah](https://forum.cursor.com/u/Zachariah)\ **Post date:** [September 22, 2026, 10:36pm UTC](https://forum.cursor.com/t/cursor-live-task-allow-list-does-not-include-grok-4-7/172552/10 "2026-09-22T22:36:44Z")
Thank you. This project uses only a local repo (mostly due to how the database dev work is currently setup) making Cloud workers a no-go for this project at this time. I’ll keep using my hack for now. Thank you for tracking the issue and replying back when it’s solved! ---
**Author:** ![Colin](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/colin/32/90452_2.png) [@Colin](https://forum.cursor.com/u/Colin)\ **Post date:** [October 1, 2026, 12:02am UTC](https://forum.cursor.com/t/cursor-live-task-allow-list-does-not-include-grok-4-7/172552/12 "2026-10-01T00:02:39Z")
Hey @Zachariah, Following up as promised. This has been addressed on our side and will be rolled out soon if it hasn’t been already. No Cursor update is needed. Subagents launched from a local Grok 4.7 agent (including ones that inherit the parent’s model) should now run on Grok 4.7 instead of silently falling back to Auto. Could you give it another try with a local Grok 4.7 agent and let me know if the workers still end up on Auto? If you pin a specific reasoning level for a worker and it doesn’t stick, let me know here too. ---
**Author:** ![vladtech](https://avatars.discourse-cdn.com/v4/letter/v/bc79bd/32.png) [@vladtech](https://forum.cursor.com/u/vladtech)\ **Post date:** [October 16, 2026, 7:05pm UTC](https://forum.cursor.com/t/cursor-live-task-allow-list-does-not-include-grok-4-7/172552/13 "2026-10-16T19:05:30Z")
@Colin — quick follow-up on pinned reasoning level (as you asked). Cursor 3.25.6, Windows, local Task: - model: “grok-4.7-xhigh” → Invalid model selection (slug not on allowlist) - model: “grok-4.7-high” → works So Grok 4.7 subagents appear available at high, but xhigh still can’t be pinned. Please add grok-4.7-xhigh to the live Task allowlist (or confirm if a different slug should be used). ---
**Author:** ![Colin](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/colin/32/90452_2.png) [@Colin](https://forum.cursor.com/u/Colin)\ **Post date:** [October 17, 2026, 11:17am UTC](https://forum.cursor.com/t/cursor-live-task-allow-list-does-not-include-grok-4-7/172552/18 "2026-10-17T11:17:50Z")
Hey @vladtech, thanks for testing that. You’re using the right slug. Right now the Task tool only offers one reasoning level per model, and it’s whichever level (and Fast setting) is currently selected for that model in your model picker. That’s why you see grok-4.7-high and not grok-4.7-xhigh. If you switch the picker to Extra High, xhigh becomes available but high drops off the list. Being able to pick any reasoning level per worker is a known limitation we’re tracking, and I’ve added your report to it.