AgentRoom：CRDT支援之共享工作區中的並發多代理編碼

摘要

並發多智能體編碼承諾在模組間實現分工、透過冗餘提供穩健性，並在多檔案專案的自然粒度上進行平行探索。即時協作編輯協定透過無衝突複製資料型別（CRDTs）為人類團隊解決了這個協調問題，但底層的 LLM 一次只產生一個 token，現有的多智能體編碼系統繼承了這種序列限制：它們要麼透過階段交接依序安排智能體，要麼在沒有協調的情況下匯集獨立樣本，而單一智能體會以單一檔案的存根並退出（stub-and-exit）的方式放棄多達一半的困難任務。AgentRoom 是一種用於並發編碼智能體的即時協作編輯協定。其執行時層在 CRDT 合併的共用檔案系統上，將檔案層級的聲明、狀態和廣播以 MCP 工具的形式暴露出來。五個前沿的編碼-CLI 模型執行了四個後端編碼任務，並在 Python DevBench 和 Rust+axum 中進行跨語言檢查。對於 CLI 穩定的模型，配備 2 個智能體的 AgentRoom 比 Solo 放棄更少的任務，且執行間變異更小。在匹配的計算資源下，一項平均為正的 LLM 評判對比使 AgentRoom 勝過平行合併（parallel-merge）。另一個對比，即捆綁探測（bundle probe），使完整的 AgentRoom 優於每個部分案例：這是一種排序而非百分比的分配。承擔重擔的是協調，而非平行處理或 CRDT 合併。

English

Concurrent multi-agent coding promises division of labor across modules, robustness through redundancy, and parallel exploration at the natural granularity of multi-file projects. Realtime collaborative editing protocols solve this coordination problem for human teams via Conflict-free Replicated Data Types (CRDTs), but the LLMs underneath generate one token at a time and existing multi-agent coding systems inherit this serial limit: they either sequence agents through phase handoffs or pool independent samples without coordination, and a single agent abandons up to half of hard tasks with a one-file stub-and-exit. AgentRoom is a realtime collaborative editing protocol for concurrent coding agents. Its runtime layer exposes file-level claim, status, and broadcast as MCP tools on a CRDT-merged shared filesystem. Five frontier coding-CLI models ran four backend coding tasks, with cross-language checks in Python DevBench and Rust+axum. For CLI-stable models, AgentRoom with 2 agents abandons fewer tasks than Solo and has less run-to-run variation. At matched-compute, one positive mean LLM-judge contrast puts AgentRoom over parallel-merge. The other contrast, a bundle probe, puts full AgentRoom above each partial case: an ordering rather than a percentage split. Coordination, not parallelism or CRDT-merge, bears the load.