OthmanAdi/planning-with-files：README 來源編輯指南

OthmanAdi/planning-with-files 的 README 將專案描述為「Persistent file-based planning for AI coding agents and long-running tasks. Crash-proof markdown plans, session recovery after /clear and compaction, per-turn re-injection against context rot, deterministic completion gate. Manus-style.」。本文只整理倉庫可直接核對的內容，不把 star、Fork 或宣傳語當成品質證明。README 在「README」下寫到：Persistent file-based planning for AI coding agents and long-running agent tasks: the skill keeps taskplan.md , findings.md , and progress.md on disk and re-injects them every turn, so the plan survives context loss, /clear , crashes, and。這說明的是專案邊界，不是已完成的生產驗證。

從 README 的「The Problem」與相關條目，可以先判斷它是否處理你的實際問題：Goal drift: after 50+ tool calls, the original goals get crowded out。若需求不同，不應只因專案熱度就採用。本文保留原始專案名、命令與元件名，方便回到一手來源核對。 README 另外列出一項可核對的資訊：Volatile memory: the TodoWrite list disappears on context reset。這類原文條目可用來設計試跑步驟，但不能取代實際環境測試。

README 將運作方式分散在「Before and after /clear」等段落。可確認的線索包括：> you: continue > > agent: I don't have context from an earlier session. Can you describe the task and where you left off?。本文不把未寫出的架構、效能或安全邊界補成結論；真正的執行鏈仍要配合目錄、設定檔與版本標籤檢查。

第一次安裝應從 README 指出的入口開始。目前可核對的命令是： your-project/ ├── task_plan.md ← phases + checkboxes; the resume point after /clear ├── findings.md ← research notes and decisions, appended as you go └── progress.md ← session log and test results 如果倉庫沒有命令，本文不會自行編造步驟，而是建議先閱讀「The Problem」，確認系統依賴、預設埠與首次初始化。

日常使用取決於專案文件。README 的「Before and after /clear」段落提到：The agent re-reads the repo, asks you to restate the goal, and rediscovers work it already finished.。設定檔、環境變數、權限與資料目錄只在來源明確時才會記錄；沒有寫出的預設值，應在測試環境驗證並保留回滾副本。 同一部分也提到：Hidden errors: failures are not tracked, so the same mistakes repeat。