# Anthropic 原语在企业中的落地范式

- 来源：DAIR.AI (@dair_ai)
- 发布时间：2026-08-24 22:33
- AIHOT 分数：29
- AIHOT 链接：https://aihot.virxact.com/items/cmt7ckfh3222vro732d9abcim
- 原文链接：https://x.com/dair_ai/status/2091896571730493746

## AI 摘要

前沿模型虽降低了编写定制代码的成本，但并未降低审查与维护成本，每个定制方案都需从头阅读代码库。该立场论文主张采用“harness 范式”：一个编码智能体 harness 作为骨干，在所有部署中运行相同代码，无需修改。论文基于三项近期发现，其中 harness 选择对智能体基准结果差异的影响大于模型选择。

## 正文

// Applying Anthropic Primitives at Large Enterprises //

Frontier models collapsed the cost of writing custom code.

But they did not reduce the cost of reviewing and maintaining it.

Each bespoke solution drifts from the next. This means that there is a huge tax that goes into understanding them because this means reading their codebase from scratch.

The usual answers, an off the shelf product, a graph orchestration framework wired per use case, or a low code platform used as orchestrator, are custom every time and limited in scope.

This position paper argues for the harness paradigm instead. One coding agent harness runs unmodified as the backbone, with identical code across every deployment.

It builds on three recent findings. Harnesses suffice at the task level and outperform more elaborate architectures on enterprise work. Harness choice accounts for more variance in agent benchmark results than model choice does.

Paper: https://arxiv.org/abs/2608.20622

Track more trending AI papers in our academy: https://academy.dair.ai/