Environment:
- Antigravity IDE 2.5.5 (latest release)
- macOS 15.6.1 (24G90)
- MacBook Air M4, Mac16,13
Antigravity IDE immediately creates a macOS NoIdleSleepAssertion and
retains it indefinitely, even when no agent task is running.
Steps to reproduce:
- Ensure Antigravity IDE is stopped.
- Run: pmset -g assertions
- Start Antigravity IDE.
- Run the command again.
Actual result:
The Electron parent process owns:
NoIdleSleepAssertion named: "Electron"
The assertion remains for the entire application lifetime. In one
observation it persisted for 94:57:31, then was released immediately
when the IDE exited:
PID 38654(Electron) Released NoIdleSleepAssertion "Electron"
94:57:31
The behavior reproduces after relaunch: the assertion appears within
seconds and disappears when the IDE exits.
Expected result:
Antigravity IDE should not prevent system idle sleep merely because the
application is open. If a running agent operation requires a power
assertion, it should be scoped to that operation and released afterward.
Impact:
The Mac cannot enter ordinary automatic idle sleep while Antigravity IDE
is running, potentially causing battery drain and heat when users expect
the machine to sleep.
This report intentionally does not attribute separate Bluetooth dark-wake
events to Antigravity; the persistent NoIdleSleepAssertion itself is
directly reproducible.
Hi @vadim0x60,
Welcome to the Forum,
Thanks for reporting this! I tried reproducing this behavior locally, but I wasn’t seeing any sleep prevention assertions being registered by Antigravity in my tests.
To help me isolate whether Antigravity is the active culprit holding a power assertion on your Mac, could you capture some quick diagnostic info for me?
Steps to capture diagnostics:
-
Make sure Antigravity is running and your workspace is open.
-
Open your macOS Terminal application.
-
Run the following command:
pmset -g assertions
-
Copy and paste the entire terminal output back into this thread.
This will tell us if the issue is a genuine wake-lock leak inside Electron wrapper or if another application (like a browser extension) is keeping your macOS machine awake.
Having same issue.
PID 28443 is/was Antigravity
2026-09-07 01:44:08 -0700
Assertion status system-wide:
BackgroundTask 0
ApplePushServiceTask 0
UserIsActive 1
PreventUserIdleDisplaySleep 0
SoftwareUpdateTask 0
PreventSystemSleep 0
ExternalMedia 0
InternalPreventDisplaySleep 1
PreventUserIdleSystemSleep 1
NetworkClientActive 0
Listed by owning process:
pid 486(backupd): [0x0003c6a30001a576] 00:56:24 PreventUserIdleSystemSleep named: "Backup Job"
pid 59149(mdsync): [0x0003c71a000ba595] 00:54:25 BackgroundTask named: "com.apple.metadata.mds.scan"
pid 28443(Electron): [0x0003a8fa000198ee] 03:02:57 NoIdleSleepAssertion named: "Electron"
pid 415(WindowServer): [0x0003ca660009a5d8] 00:00:00 UserIsActive named: "com.apple.iohideventsystem.queue.tickle serviceID:100000c9b service:AppleHIDKeyboardEventDriverV2 product:Apple Internal Keyboard / Trackpad eventType:3"
Timeout will fire in 1800 secs Action=TimeoutActionRelease
pid 354(powerd): [0x0003cdfe0001a622] 00:25:01 PreventUserIdleSystemSleep named: "Powerd - Prevent sleep while display is on"
pid 354(powerd): [0x0003d3940010834c] 00:01:11 InternalPreventDisplaySleep named: "com.apple.powermanagement.delayDisplayOff"
Timeout will fire in 229 secs Action=TimeoutActionTurnOff
No kernel assertions.
It does the same thing on Windows. V 2.14.0. I’m fine with it keeping the system on when i have prevent sleep enabled because i access agy remotely often but that should not prevent the monitor from sleeping.