Wechaty - the missing bot SDK for WeChat

本文探讨了Wechaty如何成为微信聊天机器人开发的重要工具,尤其是在缺乏官方SDK的情况下。通过Wechaty,开发者可以将个人微信账号转变为聊天机器人,并利用其提供的JavaScript API进行交互。文章还讨论了群聊API、多用户聊天机器人以及机器人间通信的可能性。

Recently I was in San Francisco talking to a lot of my friends in the Chatbot community, and realized how the Wechaty community is becoming central to Chatbot development in China.

A lot of Americans look to China and WeChat for the future of Chatbots, but in fact Tencent has still not released an official “Chatbot SDK”. Unlike Facebook, Slack, LINE, Telegram and every other messaging platform. bot.QQ.com is still just a teaser of what might be.

If you are determined enough, you can still build chatbots for WeChat but it feels like swimming upstream without a guide. For those who don’t know, Wechaty allows you to turn a wechat personal account into a chatbot, and provides a javascript API to talk to WeChat.

So in many ways Wechaty is becoming the missing bot SDK for WeChat, and it’s community is the main gathering point for all types of chatbot and NLP discussion in China.

Additionally the core wechaty team make a big effort to use English in their documents and community. So this community is the gateway to China for international chatbot companies too. It’s great to get some ideas from outside of the genepool too XD

Groups APIs and multiuser chatbots

So far the Tencent official APIs have only addressed 1:1 chatbots. Wechaty API is addressing an extra layer with addressing for users within groups. In the US Slack is providing new APIs for botmakers, such as “buttons” that you can use to make choices in a shared chat. This brings up surprising design problems such as - if one person pushes a button to change state, should everyone else in the group see that state reflected? Or should each person get the chance to make their own decision, such as for a voting app, where everyone can vote differently. Slack is providing infrastructure to support both use cases. The official tencent API allows for messages to single users, but Wechaty also has a ‘room’ concept to allow the multiuser messaging.

With a group shared experience, there is a huge benefit to chatbots. Suddenly you can see what other people are doing. You can learn how to “use” the bot just by watching someone else’s conversation. You can do tasks together such as peer-learning. The simple text UI suddenly becomes a huge benefit as it’s easy to update a shared state. With Wechaty’s ability to provide an API to a personal account, in many ways it’s showing the way ahead for Wechat and how chat platforms can be a huge benefit to users over conventional websites and apps.

Bot to Bot communications

This talk of groups and multiuser communication reminds me of a meeting last week in SF with Beerud from Gupshup. As well as being a major player in the bot authoring space, they also run lots of infrastructure for things like telco messaging. He has a vision called “interbot” which is about bot to bot communication channels

Of course this won’t replace APIs but sometimes for a small business like a restaurant, a “chatbot” maybe the only API they’ll provide to their menu, or making a booking for example. Gupshup is developing wrappers to allow you to compose new bots by configuring and routing messages. This is a little bit similar to some of the bots people have made with Wechaty that for example route messages between different chat groups.

They compare the Bot-to-bot vs traditional APIs:

Bot to Bot

  • Connect bots without coding skills
  • Bots are generally flexible and fault tolerant
  • Bots are asynchronous, enable more possibilities
  • Bots can do bi-directional messaging
  • Bots can talk to bots and humans alike

APIs

  • Integrating APIs requires coding skills
  • APIs are fragile; they break with input errors
  • APIs are synchronous; they don’t like delays
  • APIs are uni-directional
  • APIs are meant for computers alone

There are some fascinating ideas here for a language to compose bots

This operator is used to compose a composite bot. User's can assign a botname to a concatenated set of bots.
Usage: @C = @A | @B
Here, @C is a composite bot of 2 bots A & B concatenated serially.

Analytics for Bots

Dennis and Socks

One of those from overseas who works with the Wechaty team is Dennis Yang from Dashbot.io, one of the leading analytics providers purely for ‘bots. The ‘bot industry outside of China is at this established stage where the ecosystem is growing enough to have all types of companies from authoring to analytics to AI.

Dennis is planning to support the Wechaty API to make it easy for developers to immediately get great analytics for their wechat and wechaty bots. Ping them if you would like to get advance access!

Dennis gave me some natty dashbot socks to share with the Wechaty dev team :)

The future of Wechaty

Previously I’ve seen unofficial APIs be embraced by forward looking platforms. For example this blog post from Discord

The Robot Revolution has Unofficially Begun (Unofficial API)

… a dedicated group of ballers have reverse engineered our API to create what we’re calling the official unofficial Discord API …

I believe the Telegram API also began as a reverse engineering project, that gave inspiration to the telegram team to embrace and create their own bot platform.

Let’s hope that Tencent will look to Wechaty and enable more functions for chatbot developers, or even provide a full Chatbot SDK!

Join Us

Join the group to chat with us on wechat!

Wechaty is used in many ChatBot projects by hundreds of developers. If you want to talk with other developers, just scan the following QR Code in WeChat with secret code wechaty, join our Wechaty Developers’ Home now.

Wechaty Developers’ Home

shanghai wechaty peeps

Scan now, because other Wechaty developers want to talk with you too! (secret code: wechaty)

AI 代码审查Review工具 是一个旨在自动化代码审查流程的工具。它通过集成版本控制系统(如 GitHub 和 GitLab)的 Webhook,利用大型语言模型(LLM)对代码变更进行分析,并将审查意见反馈到相应的 Pull Request 或 Merge Request 中。此外,它还支持将审查结果通知到企业微信等通讯工具。 一个基于 LLM 的自动化代码审查助手。通过 GitHub/GitLab Webhook 监听 PR/MR 变更,调用 AI 分析代码,并将审查意见自动评论到 PR/MR,同时支持多种通知渠道。 主要功能 多平台支持: 集成 GitHub 和 GitLab Webhook,监听 Pull Request / Merge Request 事件。 智能审查模式: 详细审查 (/github_webhook, /gitlab_webhook): AI 对每个变更文件进行分析,旨在找出具体问题。审查意见会以结构化的形式(例如,定位到特定代码行、问题分类、严重程度、分析和建议)逐条评论到 PR/MR。AI 模型会输出 JSON 格式的分析结果,系统再将其转换为多条独立的评论。 通用审查 (/github_webhook_general, /gitlab_webhook_general): AI 对每个变更文件进行整体性分析,并为每个文件生成一个 Markdown 格式的总结性评论。 自动化流程: 自动将 AI 审查意见(详细模式下为多条,通用模式下为每个文件一条)发布到 PR/MR。 在所有文件审查完毕后,自动在 PR/MR 中发布一条总结性评论。 即便 AI 未发现任何值得报告的问题,也会发布相应的友好提示和总结评论。 异步处理审查任务,快速响应 Webhook。 通过 Redis 防止对同一 Commit 的重复审查。 灵活配置: 通过环境变量设置基
【直流微电网】径向直流微电网的状态空间建模与线性化:一种耦合DC-DC变换器状态空间平均模型的方法 (Matlab代码实现)内容概要:本文介绍了径向直流微电网的状态空间建模与线性化方法,重点提出了一种基于耦合DC-DC变换器的状态空间平均模型的建模策略。该方法通过数学建模手段对直流微电网系统进行精确的状态空间描述,并对其进行线性化处理,以便于系统稳定性分析与控制器设计。文中结合Matlab代码实现,展示了建模与仿真过程,有助于研究人员理解和复现相关技术,推动直流微电网系统的动态性能研究与工程应用。; 适合人群:具备电力电子、电力系统或自动化等相关背景,熟悉Matlab/Simulink仿真工具,从事新能源、微电网或智能电网研究的研究生、科研人员及工程技术人员。; 使用场景及目标:①掌握直流微电网的动态建模方法;②学习DC-DC变换器在耦合条件下的状态空间平均建模技巧;③实现系统的线性化分析并支持后续控制器设计(如电压稳定控制、功率分配等);④为科研论文撰写、项目仿真验证提供技术支持与代码参考。; 阅读建议:建议读者结合Matlab代码逐步实践建模流程,重点关注状态变量选取、平均化处理和线性化推导过程,同时可扩展应用于更复杂的直流微电网拓扑结构中,提升系统分析与设计能力。
内容概要:本文介绍了基于物PINN驱动的三维声波波动方程求解(Matlab代码实现)理信息神经网络(PINN)求解三维声波波动方程的Matlab代码实现方法,展示了如何利用PINN技术在无需大量标注数据的情况下,结合物理定律约束进行偏微分方程的数值求解。该方法将神经网络与物理方程深度融合,适用于复杂波动问题的建模与仿真,并提供了完整的Matlab实现方案,便于科研人员理解和复现。此外,文档还列举了多个相关科研方向和技术服务内容,涵盖智能优化算法、机器学习、信号处理、电力系统等多个领域,突出其在科研仿真中的广泛应用价值。; 适合人群:具备一定数学建模基础和Matlab编程能力的研究生、科研人员及工程技术人员,尤其适合从事计算物理、声学仿真、偏微分方程数值解等相关领域的研究人员; 使用场景及目标:①学习并掌握PINN在求解三维声波波动方程中的应用原理与实现方式;②拓展至其他物理系统的建模与仿真,如电磁场、热传导、流体力学等问题;③为科研项目提供可复用的代码框架和技术支持参考; 阅读建议:建议读者结合文中提供的网盘资源下载完整代码,按照目录顺序逐步学习,重点关注PINN网络结构设计、损失函数构建及物理边界条件的嵌入方法,同时可借鉴其他案例提升综合仿真能力。
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值