Methodology
Evidence before publication
Codex100 is an independent tracker for operator-reviewed public Codex quota resets. It distinguishes confirmed official evidence from observed resets and never reads private accounts.
Discovery
Public messages are collected as review candidates. Collection and automatic classification never claim that a reset has happened.
Human confirmation
An operator verifies the public source, occurrence time, scope and reset type before creating a published event. Corrections retain an audit trail.
Published types
Global reset, Banked Reset and quota adjustment. Community +1 requests are displayed separately and never count as evidence.
Sources and citation
Every published record links to its canonical public evidence. When citing a reset, use its occurrence time, reset type and evidence status. Link the reviewed public source and do not present an observed reset as an official announcement.
Independence
Codex100 is not affiliated with, endorsed by, or operated by OpenAI. It does not access private Codex accounts.