LogoThread Easy
  • Explorar
  • Criar thread
LogoThread Easy

Seu parceiro completo para threads do Twitter

© 2025 Thread Easy All Rights Reserved.

Explorar

Newest first — browse tweet threads

Keep on to blur preview images; turn off to show them clearly

Don't know how to start your entrepreneurial journey?

Consider my behind-the-scenes, 111-lesson course called The Creator MBA.

How to build, launch, and scale a lean, profitable, one-person business.

75% off for early Black Friday: https://t.co/GEcIgpp6Cj

Don't know how to start your entrepreneurial journey? Consider my behind-the-scenes, 111-lesson course called The Creator MBA. How to build, launch, and scale a lean, profitable, one-person business. 75% off for early Black Friday: https://t.co/GEcIgpp6Cj

The $10M Solopreneur | Helping 100,000+ burned-out corporate professionals build six-figure, one-person online businesses at https://t.co/27OAdtwjh5

avatar for Justin Welsh
Justin Welsh
Thu Nov 27 00:20:10
Regarding YouTube AI moderation issue:

Guys... I see all your messages, and replies, and DMs.

I'm just a lowly pawn in their empire... there's not much I can do.

But I have been messaging a few of my contacts there to try to raise the issue.

Hopefully we can create change.

Regarding YouTube AI moderation issue: Guys... I see all your messages, and replies, and DMs. I'm just a lowly pawn in their empire... there's not much I can do. But I have been messaging a few of my contacts there to try to raise the issue. Hopefully we can create change.

https://t.co/zSf5Z2H78P https://t.co/ryMAyS77qn https://t.co/Gm6gdHaLgp On a mission to inspire 1B people to build stuff!

avatar for Pat Walls
Pat Walls
Thu Nov 27 00:19:02
Tbh I think Tsinghua, PKU and others should just expand, as it stands they're subsidizing top American/EU universities (which also should expand) by making them unreasonably attractive

Tbh I think Tsinghua, PKU and others should just expand, as it stands they're subsidizing top American/EU universities (which also should expand) by making them unreasonably attractive

We're in a race. It's not USA vs China but humans and AGIs vs ape power centralization. @deepseek_ai stan #1, 2023–Deep Time «C’est la guerre.» ®1

avatar for Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)
Thu Nov 27 00:16:02
Cursor 中构建功能与修复 bug 的实用流程

来自 @ericzakariasson 的分享,详细阐述了他使用 Cursor 进行软件开发的具体工作流。流程强调人类与 AI 的协作:AI 负责规划和生成代码,而开发者全程把关,确保代码质量。它突出“始终审查代码”的核心原则,避免 AI 过度自动化带来的风险。

核心步骤
Eric 的流程是一个循环过程,从理解问题入手,到代码生成、验证和优化。

1. 预热理解(使用高级模型):  
   选择 Claude Opus 4.5 或 GPT-5.1-Codex 等更智能的模型,向其查询“即将进行的变更”。目的是快速把握相关代码的上下文,例如重构限流逻辑时,先确认涉及的文件和依赖。这一步帮助开发者设定预期,避免盲目规划。

2. 切换规划模式:  
   进入 Cursor 的“计划模式”,用相同模型描述变更需求。AI 会生成一个 Markdown 格式的详细计划,包括步骤、潜在影响和备选方案。计划可编辑,便于微调。

3. 互动迭代: 
   回复 AI 的跟进问题(如澄清边缘案例或优先级),逐步完善计划。这体现了“智能体”的协作性:它不是被动工具,而是主动提问以提升准确度。

4. 生成与初步审查:  
   使用 Composer-1 基于计划输出代码。同时,进行轻量审查:检查逻辑一致性、潜在 bug 和性能瓶颈。

5. 人工验证:  
   开发者亲自运行测试,确认解决方案在实际环境中有效。这一步强调人类判断的不可或缺性。

6. 后续调整:  
   如需进一步变更,@提及原有计划,切换回 Composer-1 的计划模式迭代。整个 2-5 步可循环,直至满意。

7. 深度审查与重构:  
   若 Composer-1 不足以解决问题,退回 Claude Opus 4.5/GPT-5.1-Codex 进行彻底代码审计。然后,重构优化:复用现有代码,或采用更高效的实现方式。最终,再次全盘审查。

关键洞见
· 模型混合使用:前期用 SOTA 模型理解复杂性,后期用 Composer-1 生成高效代码。这种“最佳工具适配”策略最大化了 AI 的优势,同时控制成本。
· 审查为王:帖子反复强调“always review the code”,因为 AI 生成虽快,但易忽略上下文或引入隐蔽错误。相比传统 CLI 编码智能体,Cursor 的可视化审查更高效。
· 适用扩展:适用于功能开发、bug 修复,甚至全软件开发生命周期。Bugbot 等工具可辅助,但核心仍是此循环。

Cursor 中构建功能与修复 bug 的实用流程 来自 @ericzakariasson 的分享,详细阐述了他使用 Cursor 进行软件开发的具体工作流。流程强调人类与 AI 的协作:AI 负责规划和生成代码,而开发者全程把关,确保代码质量。它突出“始终审查代码”的核心原则,避免 AI 过度自动化带来的风险。 核心步骤 Eric 的流程是一个循环过程,从理解问题入手,到代码生成、验证和优化。 1. 预热理解(使用高级模型): 选择 Claude Opus 4.5 或 GPT-5.1-Codex 等更智能的模型,向其查询“即将进行的变更”。目的是快速把握相关代码的上下文,例如重构限流逻辑时,先确认涉及的文件和依赖。这一步帮助开发者设定预期,避免盲目规划。 2. 切换规划模式: 进入 Cursor 的“计划模式”,用相同模型描述变更需求。AI 会生成一个 Markdown 格式的详细计划,包括步骤、潜在影响和备选方案。计划可编辑,便于微调。 3. 互动迭代: 回复 AI 的跟进问题(如澄清边缘案例或优先级),逐步完善计划。这体现了“智能体”的协作性:它不是被动工具,而是主动提问以提升准确度。 4. 生成与初步审查: 使用 Composer-1 基于计划输出代码。同时,进行轻量审查:检查逻辑一致性、潜在 bug 和性能瓶颈。 5. 人工验证: 开发者亲自运行测试,确认解决方案在实际环境中有效。这一步强调人类判断的不可或缺性。 6. 后续调整: 如需进一步变更,@提及原有计划,切换回 Composer-1 的计划模式迭代。整个 2-5 步可循环,直至满意。 7. 深度审查与重构: 若 Composer-1 不足以解决问题,退回 Claude Opus 4.5/GPT-5.1-Codex 进行彻底代码审计。然后,重构优化:复用现有代码,或采用更高效的实现方式。最终,再次全盘审查。 关键洞见 · 模型混合使用:前期用 SOTA 模型理解复杂性,后期用 Composer-1 生成高效代码。这种“最佳工具适配”策略最大化了 AI 的优势,同时控制成本。 · 审查为王:帖子反复强调“always review the code”,因为 AI 生成虽快,但易忽略上下文或引入隐蔽错误。相比传统 CLI 编码智能体,Cursor 的可视化审查更高效。 · 适用扩展:适用于功能开发、bug 修复,甚至全软件开发生命周期。Bugbot 等工具可辅助,但核心仍是此循环。

邵猛,中年失业程序员 😂 专注 - Context Engineering, AI Agents. 分享 - AI papers, apps and OSS. ex Microsoft MVP 合作 - 私信/邮箱:shaomeng@outlook.com 📢 公众号/小红书: AI 启蒙小伙伴

avatar for meng shao
meng shao
Thu Nov 27 00:14:06
RT @BeidiChen: 📘 Holiday read! From Software Engineer to AI Environment Architect

🚀 Tldr of our blog: We see an exciting future where engi…

RT @BeidiChen: 📘 Holiday read! From Software Engineer to AI Environment Architect 🚀 Tldr of our blog: We see an exciting future where engi…

Cartoonist, Engineer, PM, Partner @a16z investing in infra & AI Prev Product lead @HashiCorp, Founding Eng/PM @Transposit. Eng @AppDynamics. Opinions = own.

avatar for Yoko
Yoko
Thu Nov 27 00:11:59
RT @zarazhangrui: A student asked me for career advice. I uploaded the transcript into @Replit Design and asked it to turn my ideas into a…

RT @zarazhangrui: A student asked me for career advice. I uploaded the transcript into @Replit Design and asked it to turn my ideas into a…

ceo @replit. civilizationist

avatar for Amjad Masad
Amjad Masad
Thu Nov 27 00:07:30
  • Previous
  • 1
  • More pages
  • 2252
  • 2253
  • 2254
  • More pages
  • 5634
  • Next