今日速览
| 序号 | 标题 | 来源 | 日期 | 主题 | 推荐等级 |
|---|---|---|---|---|---|
| 1 | tap: A File-Based Protocol for Heterogeneous LLM Agent Collaboration | arXiv | 2026-06-12 | AI-Agent | 高 |
| 2 | Communication Policy Evolution for Proactive LLM Agents | arXiv | 2026-06-12 | AI-Agent | 高 |
重点论文与技术动态
1. tap: A File-Based Protocol for Heterogeneous LLM Agent Collaboration
- 来源：arXiv
- 日期：2026-06-12
- 作者/机构：Minseo Kim
- 主题标签：
AI-Agent,arXiv - 推荐等级：高
- 分类：cs.SE, cs.AI, cs.HC
一句话结论
tap协议通过文件优先设计实现了Claude (Anthropic)和Codex (OpenAI)等异构LLM代理在共享代码库上的无缝协作，无需共享内存或相同运行时，27天自应用实验验证其能显著提升代码审查质量。
核心内容
- tap协议允许不同供应商LLM代理（如Claude和Codex）直接协作，无需共享内存或相同运行时环境。
- 核心采用文件优先设计，消息文件带元数据保留为原始消息，结合文件通信（Tier 1）和实时通知（Tier 2），工作通过git工作树隔离。
- 实验分析375个审查工件显示，异构模型对缺陷率69.8%显著高于同构模型53.1%，证明协作拓宽审查视角。
方法与数据
- 方法：文件优先设计，消息文件带元数据，结合文件通信（Tier 1）和实时通信（Tier 2）。
- 数据：27天37代自应用操作收集209个pull请求和717个操作工件。
价值判断
- 值得关注：异构模型协作显著提升缺陷检测率（69.8% vs 53.1%），验证跨供应商协作优势。
- 可复用点：文件消息保存机制可推广至其他LLM代理协作系统。
- 局限/待核查：摘要未明确。
摘要
Existing multi-agent software development systems have proposed many forms of agent collaboration, including role-based collaboration and automated code review. However, many systems assume a common runtime, a central conversation server, or the same API family. Under these assumptions, LLM agents from different vendors cannot easily exchange messages directly from their own execution environments while dividing development and review work on a shared codebase. This paper presents tap, a file-based collaboration protocol that allows Claude (Anthropic) and Codex (OpenAI) to collaborate on one codebase without shared memory or an identical runtime. The core of tap is a file-first design that preserves markdown files with metadata as original messages, combines a file inspection path (file communication, Tier 1) with real-time notification paths for Claude and Codex (real-time communication, Tier 2), and isolates work through separate git worktrees. Even if real-time notification fails or a receiver restarts, the message file remains available and the same content can be inspected again. In a 27-day, 37-generation self-applied operation where tap was used to develop and review itself, we collected 209 tap-related pull requests and 717 operational artifacts. An analysis of 375 review artifacts showed that the share of reviews recording at least one defect or requested change was 69.8% for heterogeneous model pairs and 53.1% for homogeneous model pairs. These results show that tap, which combines file-based message preservation with real-time notification, operates in a real production repository, and that combining heterogeneous models and execution environments can broaden review perspectives. tap is distributed as the open-source npm package @hua-labs/tap (v0.5.2).2. Communication Policy Evolution for Proactive LLM Agents
- 来源：arXiv
- 日期：2026-06-12
- 作者/机构：Xinbei Ma, Jiyang Qiu, Yao Yao, Zheng Wu, Yijie Lu, Xiangmou Qu 等 13 人
- 主题标签：
AI-Agent,arXiv - 推荐等级：高
- 分类：cs.AI
一句话结论
该研究提出通信策略进化（CPE）框架，通过融合文本与UI交互优势，显著提升LLM代理任务成功率，揭示通信行为为关键设计维度。
核心内容
- 形式化通信策略，建立文本和UI基础策略，并在多样环境、角色及模型组合中系统评估其效果。
- 实验表明文本交互有效提升任务性能，而结构化UI交互则改善响应质量与角色合规性。
- CPE框架通过rollout和提示级进化实现通信策略自优化，无需模型修改即可在多设置中达成最佳任务成功。
方法与数据
- 实验设置包括User-Agent和Planner-Executor两个互补场景，用于构建信息不对称。
- 摘要未明确具体数据集或实验细节。
价值判断
- 值得关注：通信行为作为LLM代理的关键设计维度，长期被忽视，本研究首次系统探索其影响。
- 可复用点：CPE框架仅依赖提示精炼优化策略，无需模型改动，可直接复用于其他LLM代理系统。
- 局限/待核查：摘要未提及研究局限或待验证点。