---
id: 20260805-T0-05
title: "首份生产规模Copilot追踪报告：Agent编码工作负载特征曝光"
title_en: "First Production-Scale Study Reveals Copilot Agent Coding Traces"
url: https://ai.daily.yangsir.net/daily/20260805-T0-05
issue_date: 2026-08-05
publish_date: 2026-08-04T04:00:00.000Z
category: research
source_name: "arXiv cs.AI"
source_url: https://arxiv.org/abs/2608.00101
---

# 首份生产规模Copilot追踪报告：Agent编码工作负载特征曝光

arXiv新论文首次以生产规模数据呈现AI编码Agent的工作负载特征，涵盖GitHub Copilot、Claude Code和Codex。研究发现，这些Agent将多步LLM推理与工具执行交错进行，形成与聊天机器人截然不同的工作负载模式。研究人员对采样追踪数据进行了深入特征分析，包括推理调用频率、工具使用序列等，为优化Agent性能和系统设计提供了数据基础。

## English Version

**First Production-Scale Study Reveals Copilot Agent Coding Traces**

A new arXiv paper presents the first production-scale characterization of AI coding agents like GitHub Copilot, Claude Code, and Codex. It finds these agents interleave multi-step LLM inference with tool execution, creating a workload distinct from chatbots. Detailed analysis covers inference call frequency and tool sequences, providing data-driven insights for optimizing agent performance and system design.

---

**来源**：[arXiv cs.AI](https://arxiv.org/abs/2608.00101)

**详情页**：https://ai.daily.yangsir.net/daily/20260805-T0-05

---

*智语观潮 · Daily — https://ai.daily.yangsir.net/llms.txt*