与「harness」相关的搜索结果

其实让Codex拥有Hermes一样高效工作。 只需要装个oh-my-openagent 开源项目。 它不是普通插件。 而是给 AI 编程工作流加一层 Agent Harness。 自动读取项目规则、接入 LSP / CodeGraph、 检查代码注释、支持持续执行/续跑、子 Agent 协作。 这样Codex升级成“有规则、有角色、有检查、有沉淀”的工程团队。 开源链接评论区⬇️

好奇有没有一些小团队在做 “Harness Engineering as a Service” 的事情，把自己“部署”到其他中型大型公司，基于他们已有的工作流和正在使用的各种工具，搭建一套全公司可以无痛上手的新的 Workflow/Harness

Grok 4.5 已在 SpaceX 和 Tesla 进入私有 beta 测试 基于 xAI 1.5 万亿参数 V9 基础模型，额外加入 Cursor 代码训练数据。早期评估显示性能接近甚至可能超过 Claude Opus。RL 仍在持续显著提升模型，Grok Build harness 正针对真实 SpaceX/Tesla 工程环境进行迭代优化。 Elon 提到，一批顶尖 Starlink/Starship 工程师正把大量时间转向 AI，同时 SpaceXAI 今年每月将发布全新从头训练的模型。

To be clear, I’m not saying the Grok v9 foundation model will be mind-blowingly better than anything, but it will be a solid workhorse in the same league as Opus. And the SpaceXAI cadence of model and harness improvement is speeding up tremendously, particularly due to a few dozen of the top Starlink/Starship engineers shifting a much of their time to AI. Worth noting that the v8 foundation (Grok 4.3) model is a 0.5T model that finished training all the way back in December and has many fundamental flaws, so Grok 4.5 will feel like a gigantic upgrade!

BREAKING: Elon Musk confirms Grok 4.5 is now in private beta at SpaceX and Tesla. • Early evals show performance close to, possibly exceeding Opus • Based on xAI’s 1.5T V9 foundation model • Trained with Cursor data added • Grok Build harness is getting better every day • New models trained from scratch will be released every month this year The pace at SpaceXAI is absolutely insane.

Grok 4.5, based on our 1.5T V9 foundation model, with Cursor data added in supplemental training, is now in private beta at SpaceX & Tesla. Early evals show performance close to, perhaps exceeding Opus. RL is continuing to significantly improve the model, and the Grok Build harness gets better every day. Nice work by all those involved! Completely trained from scratch new models will be released by @SpaceX every month this year.

如何把一篇口播稿👉高质量动画演示？ 市面上的HTML Skill大多只追求美观 就像漂亮的花瓶，好看但是空心的 好的口播稿的信息密度是很大的 内容需要层层推进、重点需要突出 传达才能准确，动画设计上必须精准、可操控 综合对比几种生成方案，我选择用harness驾驭工程 来定制一个稳定生成HTM，叫Jacky-motion 项目地址： 第一步：审稿：把那些太书面的表达改成口语短句，把信息重点和视觉化的抓手都标出来； 第二步：节拍分镜：哪些是核心信息、哪些词是视觉动词、哪里需要重点突出，都会在这一步跟你确认清楚； 第三步：目前有杂志编辑、报纸纪实、苹果科技、财经演播室四种风格，选好后它就会给每个节拍设计构图方案，再跟你确认一次，然后一键输出一个单文件的 HTML； 第四步：直接播放配合口播录屏或继续下一句语音生成与视频渲染合成； 欢迎体验以及提出优化建议～

my ideal AI interface is a single never-ending chat thread. i don't want to think about the concepts of sessions, context windows, worktrees, mcp servers or really anything else. The harness should automate everything transparently.