向阳乔木

Bento是一个开源HTML PPT项目，通过博客网址自动生成PPT，支持JSON编辑和AI协作。

向阳乔木和姚老师组织的GEO公开课，到时候向阳乔木负责一些AI工具和模型的问答分享。

This article introduces the Harness framework, an AI tool designed to automatically generate paper charts through a collaborative workflow involving designers, executors, validators, and revisionists.

工程师使用AI时的心态决定结果质量，知识工作因不确定性难以被替代。

Suno's music generation skill significantly improves accuracy and user experience through 6,000+ new music style entries and Google CDP automated token refresh, with open-source deployment via npx command.

元子在直播中演示了两个工具：Cohub（人机共生平台）和Gemini提示词侧边栏工具，均提供公开链接。

NotebookLM 可用于跨国小团队沟通，通过生成播客和多语种支持实现高效对齐。

该工具通过抓取AppStore用户评价并利用LLM进行数据分析，为产品经理提供可用信息，代码已开源。

乔木小说创作 Skill 是一个开源工具，可帮助用户自动生成小说剧情、人物设定等内容。

Codex supports cross-device remote collaboration: after connecting other devices in settings, you can add remote projects when creating a project locally, enabling remote control of Codex on another computer.

A guide from GPT 5.5 Pro on Codex Goal usage reveals two techniques: use /plan to let AI ask clarifying questions and converge on a Goal; use /goal to generate executable Codex instructions, improving code generation quality and consistency.

通过调试GPT-image-2提示词，发现简洁风格减少了AI味，更适配AI Agent产品，增加了推理空间。

Vista8 demonstrates the Suno-based AI music generation skill that supports on-demand generation of any style music with $10 monthly pricing and optimized to over 6000 music style retrieval capabilities.

The GEO industry faces rampant black hat methods, and Teacher Yao and Xiaoyang Qiaomu release the 'GEO Red Book' to help SMEs avoid risks.

向阳乔木分享了一个通过终端安装脚本的命令，但未提供详细的技术内容或背景。

The blogger has added an AI chat sidebar to enable natural language commands for generating illustrations and titles, with plans to open-source it upon completion.

A developer used GPT-Image-2 model with a unified prompt 'distant mountains, nearby water, and small forest' to batch-generate style samples from 380+ artists, creating an online showcase website with open-source code to help users quickly understand artistic style characteristics and apply them to AI painting creation.

Most third-party AI chat clients currently lack audio and video upload and parsing capabilities, limiting the practical use of multimodal models.

Get Notes is considered undervalued, capable of parsing audio-video content from platforms like Xiaoyuzhou, Bilibili, Douyin, YouTube, and TikTok via URL, generating summaries, with an annual membership at just $99 during promotion, offering CLI and Skill tools for Codex integration and raw transcription output.

AI significantly lowers content production costs, shifting competitive focus to topic selection, aesthetics, trust-building, and distribution efficiency.

HTML regains relevance due to the Lindy Effect in the AI era, but its dissemination is limited in China’s WeChat ecosystem, where images, PPTs, and PDFs remain more effective for sharing.

豆包手机将于9月正式开售，但推文内容缺乏技术细节和工程价值，仅属产品发布消息。

文章为AI公益课直播预告，缺乏技术深度，主要面向大学生和职场新人。

该推文预告了AI知识科普视频系列，但未提供技术细节，内容过于浅显。

该推文宣布了一个视频生成开源项目，但缺乏技术细节和深度分析，仅提供GitHub链接。

本文是一则直播邀请，介绍飞书与Codex结合的案例分享，但缺乏技术细节。

文章讨论了GLM-5.2模型的能力均衡性及可能的数据来源，但信息碎片化，缺乏深度分析。

文章内容为社交媒体上的简短观察，缺乏技术深度和实用价值。

文章介绍了一个开源图标生成工具，但内容过于简略，缺乏技术深度和实用指导。

AI生成的HTML标题line-height设置过紧，导致换行后显示效果不佳，影响用户体验。