[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-22":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":9,"language":10,"languages":9,"totalLinesOfCode":9,"stars":11,"forks":12,"watchers":13,"openIssues":13,"contributorsCount":14,"subscribersCount":14,"size":14,"stars1d":15,"stars7d":16,"stars30d":17,"stars90d":14,"forks30d":14,"starsTrendScore":18,"compositeScore":19,"rankGlobal":9,"rankLanguage":9,"license":20,"archived":21,"fork":21,"defaultBranch":22,"hasWiki":23,"hasPages":21,"topics":24,"createdAt":9,"pushedAt":9,"updatedAt":25,"readmeContent":26,"aiSummary":27,"trendingCount":14,"starSnapshotCount":14,"syncStatus":13,"lastSyncTime":28,"discoverSource":29},22,"keep-codex-fast","vibeforge1111\u002Fkeep-codex-fast","vibeforge1111","A backup-first Codex skill for keeping local Codex state fast, clean, and recoverable.",null,"Python",1387,81,2,0,36,158,434,108,18.74,"MIT License",false,"main",true,[],"2026-06-12 02:00:06","# Keep Codex Fast\n\n![Keep Codex Fast cover](assets\u002Fkeep-codex-fast-cover.png)\n\nWhen Codex starts feeling heavy after weeks of chats, terminals, logs, worktrees, and project history, this gives you a calm way to inspect what is going on and reduce local drag.\n\nThis skill helps you organize local state without losing context.\n\nThe rule is simple:\n\n> Make handoffs first. Archive, don't delete. Apply changes only when you are ready.\n\n## Three Modes\n\n- **Inspect:** report-only, no writes.\n- **Maintain:** normal apply; backs up, archives old sessions, moves stale worktrees, rotates logs, prunes dead config, and normalizes paths. It does not trim thread title\u002Fpreview metadata.\n- **Optional repair:** only with `--apply --repair-thread-metadata-bloat`; shortens oversized SQLite display title\u002Fpreview metadata after backup. The transcript stays intact.\n\n## Who This Is For\n\nUse this if Codex has started feeling slower after heavy use, especially if you:\n\n- keep long chats around\n- resume old threads often\n- work across many repos\n- run multiple terminals or dev servers\n- want maintenance to feel safe, not scary\n\n## What It Does\n\nBy default, this skill only reports. It does not write files, create backups, move folders, or change local Codex state until you explicitly ask it to.\n\nIt helps Codex:\n\n- see which local state has grown over time\n- create handoff docs before archiving old chats\n- back up important state before applying changes\n- archive old chats instead of deleting them\n- detect pathological thread title\u002Fpreview metadata that can slow chat navigation\n- move stale worktrees out of the hot path\n- rotate large logs\n- prune dead project references\n- report heavy Node\u002Fdev processes without killing them\n\n## Quick Start\n\nAsk Codex:\n\n```text\nUse $keep-codex-fast to inspect my Codex local state and recommend a safe maintenance plan.\n```\n\nCodex should show you what it found first. Then you decide what to hand off, what to keep active, and what can be archived.\n\n## Handoffs First\n\nBefore archiving old active chats, create handoff documents for any repo\u002Fsession you may want to continue.\n\nA handoff is a small continuity note. It captures:\n\n- what you were doing\n- what changed\n- what files matter\n- what commands or checks already ran\n- what is still broken or undecided\n- what to do next\n\nThat lets you archive the heavy chat and start a fresh Codex thread from the handoff.\n\nCopy this into each active repo chat you care about:\n\n```text\nCreate a comprehensive handoff document for this repo\u002Fsession before I archive Codex history.\n\nInclude:\n- repo\u002Fpath and branch\n- current goal\n- what we already completed\n- files touched or investigated\n- commands\u002Ftests already run\n- known errors, warnings, or failing checks\n- open decisions\n- constraints, user preferences, and do-not-touch areas\n- the next 3-7 concrete steps\n\nAlso include a reactivation prompt I can paste into a fresh Codex chat so it can continue from this handoff without relying on the old chat context.\n\nSave the handoff in a sensible repo-local place like docs\u002Fcodex-handoffs\u002FYYYY-MM-DD-topic.md unless this repo already has a better handoff location.\n```\n\n## Safe Apply\n\nAfter handoffs exist for the chats you care about, use this:\n\n```text\nUse $keep-codex-fast to apply safe Codex maintenance.\n\nBefore changing anything, confirm that important active repo chats have handoff docs or do not need them.\n\nThen back up first, archive instead of deleting, move stale worktrees, rotate large logs, prune dead config references, and verify the result.\n\nIf Codex is currently running, do not mutate local state. Tell me to close Codex first.\n```\n\n## Thread Title And Preview Bloat\n\nSome Codex builds can store a full first user prompt as both the thread title and the list preview. When those fields grow into hundreds of thousands of characters, thread navigation can become sluggish even before a large chat is opened.\n\nThe script reports title and preview payload size in report mode and normal apply mode. It does not trim this metadata unless you explicitly opt in:\n\n```bash\npython scripts\u002Fkeep_codex_fast.py --apply --repair-thread-metadata-bloat\n```\n\nWith that flag, after backing up and only when Codex is not running, it trims active SQLite title\u002Fpreview metadata to bounded display values. It also appends repaired titles to `session_index.jsonl`, which matches current Codex name-update storage.\n\nThis does not remove the actual conversation transcript. The full rollout JSONL remains available unless you separately archive the session.\n\nThe repair manifest stores the old full title\u002Fpreview values so you can restore them. Keep the backup folder private, especially `thread-metadata-repairs.jsonl` and `restore-thread-metadata.py`.\n\nIf you are using the skill normally, this repair does not happen automatically. Treat it as an extra recommendation only when the report shows unusually large title\u002Fpreview metadata.\n\n## Weekly Or Biweekly Reminder\n\nRecurring maintenance should be a reminder, not an automatic apply.\n\nWhy: an automation cannot know whether you created handoffs for chats you still care about. It should inspect and remind you, but not archive, move, prune, rotate, normalize, delete, or mutate anything by itself.\n\nCopy this into Codex:\n\n```text\nUse $keep-codex-fast to create a recurring Codex maintenance reminder.\n\nSchedule it weekly if I use Codex heavily, or biweekly if that seems safer.\n\nThe reminder should:\n- run the keep-codex-fast report first\n- never pass --apply or run mutating maintenance automatically\n- never archive, move, prune, rotate, normalize, delete, or mutate local Codex state\n- remind me to create comprehensive handoff docs and reactivation prompts for active repo chats before any manual apply\n- summarize active session size, archived session size, extended path candidates, old session candidates, worktree candidates, log size, and top Node\u002Fdev processes\n- report heavy Node\u002Fdev processes without killing them\n- tell me that manual apply should only happen after I confirm handoffs exist or are not needed and Codex is closed\n```\n\n## Install\n\nAsk Codex:\n\n```text\nInstall the keep-codex-fast skill from https:\u002F\u002Fgithub.com\u002Fvibeforge1111\u002Fkeep-codex-fast\n```\n\nOr clone\u002Fcopy this folder into your Codex skills directory as `keep-codex-fast`.\n\n## Advanced: Manual Script Use\n\nMost users can stay inside Codex and use the prompts above. The script is here for people who want to run it directly.\n\nReport only. This is read-only and privacy-safe by default:\n\n```bash\npython scripts\u002Fkeep_codex_fast.py\n```\n\nShow raw thread IDs, chat titles, paths, and process paths only when you need detail:\n\n```bash\npython scripts\u002Fkeep_codex_fast.py --details\n```\n\nCreate backups only, without moving or changing local state:\n\n```bash\npython scripts\u002Fkeep_codex_fast.py --backup-only\n```\n\nBackup folders can contain private local Codex metadata. Keep them on your machine, and do not publish or share them unless you have reviewed what is inside.\n\nApply core maintenance actions. This does not trim thread title\u002Fpreview metadata:\n\n```bash\npython scripts\u002Fkeep_codex_fast.py --apply --archive-older-than-days 10 --worktree-older-than-days 7\n```\n\nOptionally repair oversized title\u002Fpreview metadata only when the report recommends it:\n\n```bash\npython scripts\u002Fkeep_codex_fast.py --apply --repair-thread-metadata-bloat\n```\n\nWait for Codex to exit before applying:\n\n```bash\npython scripts\u002Fkeep_codex_fast.py --apply --wait-for-codex-exit\n```\n\n## What Can Change\n\nThe skill can safely handle:\n\n- old non-pinned active sessions\n- stale worktrees\n- large `logs_2.sqlite*` files\n- dead\u002Ftemp project entries in `config.toml`\n- Windows `\\\\?\\C:\\...` path mismatches in local SQLite text fields\n- oversized thread title and first-message preview metadata in `state_5.sqlite`, only with `--repair-thread-metadata-bloat`\n\nIt does not permanently delete chats, logs, or worktrees. It moves them into archive folders and writes backup\u002Frestore artifacts before applying changes.\n\n## Mental Model\n\n- Chats are for execution.\n- Handoff docs are for memory.\n- Archives are for history.\n- Fresh threads are for speed.\n\n## Flow\n\n![Keep Codex Fast flow](assets\u002Fkeep-codex-fast-flow.svg)\n","Keep Codex Fast 是一个专注于保持本地 Codex 状态快速、整洁且可恢复的备份优先工具。其核心功能包括检查、维护和可选修复三种模式，能够帮助用户归档旧会话、移动过时的工作树、轮换日志、清理无效配置等，同时确保不会丢失上下文信息。该技能特别适合于长期使用 Codex 后感觉系统变慢的情况，尤其是当用户频繁地保留长对话、恢复旧线程、跨多个仓库工作或运行多个终端及开发服务器时。通过提供安全的维护方案推荐，Keep Codex Fast 让用户可以轻松管理自己的开发环境而不必担心数据丢失。","2026-06-11 02:30:30","CREATED_QUERY"]