keep-codex-fast

Codex performance degradation can be resolved by inspecting the local state and applying safe maintenance actions. The process prioritizes creating handoff documents before archiving sessions, ensuring no critical data is lost during normalization or pruning tasks.

The best way to back up and archive Codex local state safely is to use a dedicated maintenance mode. This approach preserves continuity by generating handoff documents before archiving sessions, securing your data before any normalization or pruning tasks occur.

Yes, you need Python installed to manage and inspect your Codex local state. Python is required as the core dependency, with additional dependencies needed specifically for executing backup and state normalization tasks effectively.

Codex state normalization and pruning works by safely managing local state to optimize performance. It inspects the current state, creates handoff documents to preserve session continuity, and then removes unnecessary data through archiving and pruning actions.

Yes, you can inspect your Codex local state before committing to maintenance. An Inspect mode is available to review the state and recommend a safe maintenance plan, allowing you to evaluate necessary backup and pruning actions before execution.