# Codex Chronicle generated 572 background summaries overnight and consumed quota

**URL:** <https://community.openai.com/t/codex-chronicle-generated-572-background-summaries-overnight-and-consumed-quota/1380899>\
**Category:** Codex\
**Created:** [May 14, 2026, 6:57am UTC](https://community.openai.com/t/codex-chronicle-generated-572-background-summaries-overnight-and-consumed-quota/1380899 "2026-05-14T06:57:47Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![snowywar](https://avatars.discourse-cdn.com/v4/letter/s/51bf81/32.png) [@snowywar](https://community.openai.com/u/snowywar)\
**Post date:** [May 14, 2026, 6:57am UTC](https://community.openai.com/t/codex-chronicle-generated-572-background-summaries-overnight-and-consumed-quota/1380899/1 "2026-05-14T06:57:47Z")

</div>

I am posting this here because the forum banner explicitly allows Codex developer-tool discussion, and this appears to be a Codex background-task bug.

After enabling Chronicle in the Codex desktop app, my local machine generated a large number of persisted Chronicle memory summaries during non-working hours. This appears to have consumed my 20x usage quota without explicit user turns.

Evidence summary:

- Date/time zone: 2026-05-14, Beijing time
- Chronicle resource location: ~/.codex/memories/extensions/chronicle/resources
- Total persisted Chronicle summaries generated that day: 572
- Time range: 2026-05-14 00:00:00 to 2026-05-14 11:40:00 Beijing time
- Records generated before the visible 11:30:16 Chronicle launch: 571
- From 00:00 to 09:00 Beijing time, Chronicle generated 60 summaries per hour, i.e. one persisted file per minute, even though the files are named 10min-memory-summary.

Hourly count:

```text  
60 2026-05-14 00:00  
60 2026-05-14 01:00  
60 2026-05-14 02:00  
60 2026-05-14 03:00  
60 2026-05-14 04:00  
60 2026-05-14 05:00  
60 2026-05-14 06:00  
60 2026-05-14 07:00  
60 2026-05-14 08:00  
30 2026-05-14 09:00  
2 2026-05-14 11:00  
```

Direct log evidence shows automatic summary sessions:

```text  
2026-05-14T03:30:16.739Z … codex\_chronicle starting  
2026-05-14T03:30:16.783Z … screen recording starting  
2026-05-14T03:40:36.867Z … starting summary session … level=“10min” bucket=2026-05-14T03-30-00Z  
2026-05-14T03:40:36.868Z … starting codex exec summary session … model=gpt-5.5 reasoning\_effort=medium  
2026-05-14T03:42:36.890Z … starting summary session … level=“10min” bucket=2026-05-14T03-30-00Z  
2026-05-14T03:42:36.891Z … starting codex exec summary session … model=gpt-5.5 reasoning\_effort=medium  
```

The logs also show repeated screen-recording permission failures while Chronicle was active:

```text  
ERROR screen recording reconcile failed kind=PermissionRequired state=Unsupported  
No shareable content available: Content unavailable: user denied TCC screen capture  
```

Why this looks like a bug:

1. The persisted summary files show one 10min-memory-summary per minute for long periods.
2. 571 persisted summaries existed before the visible 11:30:16 Chronicle process launch.
3. Logs show repeated automatic codex exec summary sessions using gpt-5.5.
4. Logs also show repeated screen-capture permission failures, but Chronicle still started or retried summary-related work.
5. This background behavior appears to consume quota without explicit user requests.

I have also submitted this through OpenAI Help Center and asked support to investigate and restore/refund the consumed quota. I can provide screenshots, the full exported TSV record list, and local log excerpts privately if OpenAI staff need them.

---

<div class="post-metadata">

**Author:** ![OpenAI\_Support](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/openai_support/32/561845_2.png) [@OpenAI\_Support](https://community.openai.com/u/OpenAI_Support)\
**Post date:** [September 8, 2026, 7:16pm UTC](https://community.openai.com/t/codex-chronicle-generated-572-background-summaries-overnight-and-consumed-quota/1380899/2 "2026-09-08T19:16:15Z")

</div>

Sorry for the late reply! The current Computer History feature replaces Chronicle and has different collection controls: [https://learn.chatgpt.com/docs/customization/computer-history](https://learn.chatgpt.com/docs/customization/computer-history). That doesn’t settle the quota loss you reported. Please keep the timestamps and summary-session evidence in your existing Support case for a usage review.

---

<div class="post-metadata">

**Author:** ![Mark\_G](https://avatars.discourse-cdn.com/v4/letter/m/e8c25b/32.png) [@Mark\_G](https://community.openai.com/u/Mark_G)\
**Post date:** [September 8, 2026, 7:16pm UTC](https://community.openai.com/t/codex-chronicle-generated-572-background-summaries-overnight-and-consumed-quota/1380899/3 "2026-09-08T19:16:16Z")

</div>