multica-ai/multica：README 來源編輯指南

multica-ai/multica 的 README 將專案描述為「The open-source managed agents platform. Turn coding agents into real teammates , assign tasks, track progress, compound skills.」。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「Multica」下寫到：The open-source managed agents platform. Turn coding agents into real teammates , assign tasks, track progress, compound skills.。這說明的是專案邊界，不是已完成的生產驗證。

從 README 的「Features」與相關條目，可以先判斷它是否處理你的實際問題：Squads , group agents (and humans) under a leader agent and assign work to the squad. The leader decides who should pick it up, so routing stays stable as the team grows. @FrontendTeam instead of @alice-or-bob-or-carol.。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：Agents as Teammates , assign to an agent like you'd assign to a colleague. They have profiles, show up on the board, post comments, create issues, and report blockers proactively.。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「What is Multica?」等段落。可確認的線索包括：No more copy-pasting prompts. No more babysitting runs. Your agents show up on the board, participate in conversations, and compound reusable skills over time.。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： brew install multica-ai/tap/multica 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「What is Multica?」，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「What is Multica?」段落提到：For larger teams, Squads add a stable routing layer: assign work to a group led by an agent, and the leader delegates to the right member.。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：Autonomous Execution , set it and forget it. Full task lifecycle management (enqueue, claim, start, complete/fail) with real-time progress streaming via WebSocket.。