Forumics logo Forumics
Issue 01

Back to issue

All Codex known issues

CLAIM C059 Supported 2026-11-08

GitHub issue #34260 in openai/codex, opened 2026-07-20 by RocStone, records a report that Codex Desktop 26.715.4045.0 on Windows 25H2 x64, build 26200.8875, can enter an unbounded process-cleanup loop where hundreds of taskkill.exe processes remain alive, each with a corresponding conhost.exe, that the accumulated taskkill instances repeatedly query Win32_Process, exhaust the WMI provider quota, and make the entire Windows UI sluggish or unable to launch new processes, with peak measurements of 457 taskkill.exe processes, 473 conhost.exe processes, 1,388 total processes, over 14,000 threads, over 350,000 handles, and WMI provider private memory of 720.5 MB, and that killing the accumulated taskkill.exe processes immediately reduced the total process count from 1,388 to about 565 while WMI private memory fell to about 104 MB, but taskkill.exe counts grew again from 156 to 320 in 20 seconds while the original Codex process remained running.

Verbatim Excerpts from Cited Records
“Codex Desktop on Windows can enter an unbounded process-cleanup loop where hundreds of `taskkill.exe` processes remain alive, each with a corresponding `conhost.exe`.”
— RocStone · openai/codex (issue 34260 opening body, Summary section) · 2026-07-20T03:58:11Z ·GitHub Issue ↗
“The accumulated `taskkill` instances repeatedly query `Win32_Process`, exhaust the WMI provider quota, and make the entire Windows UI sluggish or unable to launch new processes.”
— RocStone · openai/codex (issue 34260 opening body, Summary section) · 2026-07-20T03:58:11Z ·GitHub Issue ↗
“All sampled `taskkill.exe` instances had the Codex Desktop main process (`ChatGPT.exe` inside the `OpenAI.Codex` package) as their parent.”
— RocStone · openai/codex (issue 34260 opening body, Observed behavior section) · 2026-07-20T03:58:11Z ·GitHub Issue ↗
“Killing the accumulated `taskkill.exe` processes immediately reduced the total process count from 1,388 to about 565, and WMI private memory fell from 720.5 MB to about 104 MB.”
— RocStone · openai/codex (issue 34260 opening body, Observed behavior section) · 2026-07-20T03:58:11Z ·GitHub Issue ↗
Counter-record: Issue remains open with 36 comments as of 2026-11-07; the API returned all 36. Maintainer chess-oai commented on 2026-08-05 that 26.730.7989.0 substantially improves the WMI storm but taskkill floods remain, leaving the issue open to track them. Reporters reproduced it from 2026-07-20 to 2026-11-07: DukeTwoCan still saw it on 26.803.10989.0 (2026-08-13) but reported no issues after updating to 26.810.41047 (2026-08-14), and phytodoxia reported a fresh reproduction captured 2026-11-05 on 26.1029.61003 with hundreds of taskkill.exe helpers and 462 WMI cancellation events. Several reporters cleared the accumulated state only by restarting or reinstalling, which the original report notes does not fix the underlying defect. The claim documents only the original report, not a verified current state. Falsification probe: If issue 34260 is deleted without archive, if the author edits the opening body so the quoted sentences no longer match verbatim, or if the author posts a retraction of the report, this claim drops to Withdrawn. Audit verification: Verified 100% continuous verbatim substring against official GitHub REST API by scripts/verify_claim_verbatim.py.

See also: Codex vs Cursor head-to-head comparison on 9bests