Codex Desktop bug double-counts historical reasoning, triggering compaction at 54%
ghyghoo8 · ghdev · 2026-10-14
openai/codex issue #53913 documents a double-counting bug in Codex Desktop's Responses Lite mode: when x-reasoning-included is absent, the native client re-adds old encrypted reasoning items to the context total, triggering auto-compaction far too early.
The author reconciled local counters exactly: compaction fired at 54% used (255k/475k), but lastapitotaltokens (255,367) plus a client-side estimate of 193 historical reasoning items (200,147) sums to exactly 455,514, exceeding the 450,000 threshold. The estimator formula ceil(max(floor(3encodedlength/4)-650,0)/4) was reverse-checked, and two actual generation requests confirmed the findings.
More from coding & agent
- Laude Institute launches FrontierCS 2: agents must beat deleted paper contributions blind — lateinteraction · 2026-10-23
- Index Network open-sources Social Discovery Protocol: agents negotiate fit on your behalf — carsonfarmer · 2026-10-23
- Installed a formal-verification skill: it silently installed Java, MacBook runs 10 degrees hotter — zealcaiden · 2026-10-23
- OpenAI's Decisions API hits GA: near real-time routing up to 10x faster than GPT-6 Luna — OpenAIDevs · 2026-10-23
- Ruff author warns: upgrade Node/Bun now — --use-system-ca still accepts the eMudhra cert — charliermarsh · 2026-10-23
- Proof of Agent: paste one prompt and your AI mines, grades, and self-governs — adamamcbride · 2026-10-23