PulseAugur
中
实时 12:24:42
English(EN) Same Haiku 5.5 Scores 82% or 65% on MCP Servers: The Harness Decides

Claude Haiku 5.5 的性能取决于工具链,而非模型本身

Sineframe 于 2026 年 11 月 1 日进行的最新评估显示,Claude Haiku 5.5 模型在不同“工具链”或脚手架下的表现差异显著。当由 OpenCode 1.19.5 驱动时,Claude Haiku 5.5 在 21 个 MCP 服务器上获得了 82.0% 的分数,但在由 Claude Code 2.1.312 驱动时,该分数下降到 65.4%。这种差异归因于工具链如何命名、描述和优先化工具,而不是模型本身的推理能力,尤其是在内置工具与外部 MCP 服务器工具重叠时。 AI

影响 强调了工具集成和描述在代理性能中的关键作用,建议开发人员专注于工具链优化。

排序理由 评估不同脚手架条件下模型性能的研究论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 dev.to — MCP tag 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →

Claude Haiku 5.5 的性能取决于工具链,而非模型本身

本文如何被排名

Signal score
1 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
评估不同脚手架条件下模型性能的研究论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
model release, product
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
1 days old
Coverage has settled into its steady-state source set.

完整方法见我们的编辑标准。

报道来源 [1]

  1. dev.to — MCP tag TIER_1 English(EN) · devs ·

    Same Haiku 5.5 Scores 82% or 65% on MCP Servers: The Harness Decides

    <p><a class="article-body-image-wrapper" href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F75924o7knd1ncsg74naz.png"><img alt="Simple reflex ag…