10 thought-provoking points about WeChat's integration with OpenClaw

Source: Wei Xi's Guide

Yesterday morning, my WeChat Moments feed was flooded with posts about WeChat integrating OpenClaw via a plugin.

Rumors of WeChat ending the competition are resurfacing in the industry, just like when DeepSeek integrated with WeChat search years ago.

The first message I sent to my shrimp on WeChat

But what I want to say is that its impact may not be as great as we imagine.

This is more of a small, expected task that WeChat should do.

Here are ten thought-provoking points I have:

First, let's look at the characteristics of WeChat OpenClaw itself.

First of all, it's not that WeChat has launched a new shrimp, but rather that if you already have a shrimp, now the official support allows you to chat with it within WeChat.

Specifically, it exists within WeChat as a plugin. (Official documentation)

In this sense, WeChat's previous plugin architecture was very prescient, as it could relatively flexibly support many new and experimental features.

Lu Shuyan, an early member of WeChat, wrote an article titled " The Lab Culture of the WeChat Team ," which is well worth reading.

Then it supports different versions of lobsters on the market, whether they are local lobsters, cloud lobsters, modified lobsters, or knock-off lobsters.

In theory, WeChat supports any Lobster that hasn't undergone extensive modifications to the OpenClaw plugin module. In fact, someone has already created a project based on the WeChat plugin's protocol code, enabling WeChat to support any AI backend, including ClaudeCode and CodeX, not just Lobster. The project address is: https://github.com/wong2/weixin-agent-sdk (This is very important and will be analyzed later.)

The entire integration process is very simple: install the plugin, then scan the QR code with WeChat, and you're done in just 2 minutes—even simpler than Telegram.

Then let's look at some of its minor features—

1. Group chat is not supported. (This is primarily for security reasons, which will be analyzed later.)

2. Streaming output is not supported. (It seems that only Lark supports this in China, if I remember correctly.)

3. You can change your name, but not your profile picture; you can pin your profile picture to the top.

4. It seems the Mac version hasn't been updated yet; the contact "WeChat clawbot" is not visible.

6. Supports OpenClaw's forward slash shortcut command;

7. Supports file transfer;

8. When a conversation is selected, only copy, forward, quote, and delete are supported. Multi-selection, translation, efficiency improvement, search, and favorites are not supported. (Yuanbao supports all buttons; this difference is interesting.)

9. Only one shrimp can be connected at a time.

10. Poor support for Markdown formatting;

11. Forwarding other people's conversations to "WeChat Clawbot" is not supported. (This is confusing, as WeChat's Yuanbao (a feature in WeChat) supports this.)

So, frankly speaking, quite a few things have been cut, and the experience is somewhat diminished.

For example, what I can't accept is that it only supports one shrimp. I created three sub-Agents in OpenClaw, so I can't communicate with them. But in Telegram, I only need to create one bot.

Second, WeChat's actions can't really be described as fast.

Some people say that WeChat acted very quickly, but I don't think it was that fast.

We can make a simple comparison: when DeepSeek became popular a year ago, it started on January 25th (DeepSeek began to support online search + deep thinking).

Then, WeChat search integrated DeepSeek on February 17th, meaning only 20 days had passed.

This time, WeChat's official intervention, if counted from the Spring Festival, has been going on for over a month. (This comparison is a bit rough, but it at least proves that it wasn't fast.)

Of course, dynamic search and dynamic WeChat contacts are different. I understand that it is more for security reasons. Development cost is by no means a bottleneck restricting WeChat from launching.

If Brother Long is involved, we must ensure that everything goes smoothly.

There are also differing opinions within the industry. Dachein, a prominent figure on Jike, believes that WeChat has already been quite aggressive.

I personally disagree with this statement. Being a little radical is fine, but being overly conservative is what's worrying.

Apple's push into AI is a case in point; the real problem is its inability to produce anything usable. (In fact, many content creators have already compared WeChat and Apple to each other when discussing AI.)

Third, WeChat's solution is elegant.

Tencent has previously launched many different shrimp, some local and some cloud-based.

Then, their methods of integrating with WeChat are also relatively awkward: some are through mini programs, and some are through WeChat customer service messages, which are not really native support.

I initially thought this might be because the WeChat business group felt that the competition among these products hadn't completely ended, and that WeChat would only natively support whichever product had truly established itself.

Looking back now, I realize I underestimated Brother Long's vision.

Brother Long waved his hand and said, "I don't support each of you individually, but I support them all. I support not only you, but all of them on the market."

What is top-level design?

This is called top-level design.

WeChat acts only as a remote control—you send commands in WeChat, the lobster executes them on the computer (or in the cloud), and the results are sent back to WeChat, while WeChat's own data boundaries remain completely unchanged.

This is a typical example of platform thinking.

Fourth, it is highly unlikely that WeChat will add group chat functionality to OpenClaw for a long time to come.

For WeChat, the importance of security cannot be overstated.

Doubao Phone was blocked by WeChat less than two days after its release because it used system-level permissions to simulate clicks across applications.

This is also a major reason why OpenClaw was not launched in a group chat this time.

Obviously, from a technical point of view, adding group chat is a piece of cake.

My assessment is that it will likely not be added to the group chat for a long time to come.

After all, the security of group chats is very uncontrollable.

I previously wrote an article titled "Your Lobster May Be Running Naked - Starting with a Chilling Paper, " which tells several fascinating stories (accidents).

The core issue is this: once group chats are established, people will inevitably try various methods to hack your account, which will eventually lead to a series of security problems.

Therefore, WeChat will be extremely cautious.

After all, for WeChat, its 1.4 billion social network is both its core asset and its Achilles' heel.

It dared not act rashly.

Fifth, the fact that WeChat has integrated OpenClaw has not lowered the barrier to shrimp farming.

Many people think that the integration of WeChat has lowered the barrier to entry for shrimp farming.

That's wrong.

WeChat lowers the barrier to chatting with shrimp, not the barrier to shrimp farming itself.

Those who have actually raised shrimp and can generate more value from them than from a chatbot already know this—

The core bottleneck in shrimp farming is not actually what tools to use to communicate with the shrimp.

It also has a series of costs beyond operation—

For example, the model you use is crucial. If you use a model that is not intelligent enough, the experience will be worse than that of a bean bun.

For example, the settings for soul.md and user.md are crucial; if they are not set up properly, the content will be uninteresting. Another example is your proficiency with memory and skills.

There are actually many hidden costs – that is, you need to have a deep understanding of the boundaries of what an agent can do, which is a matter of principle.

Therefore, WeChat's integration with OpenClaw is certainly beneficial to many people, reducing explicit costs. But the most important thing is that you should consciously delegate more tasks to the Agent and make more native and maximum use of it.

Really treat it as a lever.

There's a saying that goes something like this—

"At this point in time, if you do a lot of things yourself, it means your practical skills aren't very good."

VI. For WeChat, this is more like a duty it should be performed.

This event is certainly valuable and meaningful for WeChat and its users, but its impact may be smaller than we imagine.

For example, OpenClaw has supported Telegram since its inception, but do you think that has a significant positive impact on Telegram?

This is actually very difficult to assess in detail.

Users' minds are still on OpenClaw itself, not on Telegram.

This should actually be understood as follows: if you don't do it, it might have a positive effect on other IMs that do.

But if you do it, how much it will improve you is questionable.

After all, you're already part of the infrastructure.

Therefore, I am more inclined to believe that WeChat's integration this time is actually just a validation of its status as a mainstream chat infrastructure.

However, this is clearly not good news for Lark, WeChat Work, and DingTalk, which have recently been keen on highlighting one-click access to lobsters.

VII. How much value can it realize, or how much value space does OpenClaw itself have?

If you know how to raise shrimp, you can do it on Lark, WeChat Work, Telegram, and QQ.

If you don't know how to raise shrimp, joining WeChat won't change that.

It's very simple. Yuanbao actually natively supports WeChat, but in reality, how often do we communicate directly with Yuanbao within WeChat?

Most people will likely still prefer using Doubao, or even be more accustomed to opening the Yuanbao app. (Of course, it's not to say it's completely useless. When I was explaining AI to the elderly in Wenquan Town, Haidian District, some of them were used to talking to Yuanbao in WeChat because WeChat is the only app they know how to use. But OpenClaw's logic is completely different.)

WeChat appears to have created a feature accessible to everyone, but the actual audience—those who have already "raised" shrimp—is likely a relatively small percentage of WeChat's total user base.

This brings us back to an old problem: people in the tech world feel like everything has changed, while ordinary users feel like nothing has happened.

A deep understanding of technology penetration takes time, and this is also very important.

The Pareto Principle (80/20 rule) always applies. Take another look at the following image—

8. This integration is a very pure tool integration and has not been integrated with other modules of WeChat.

Essentially, this integration simply adds a separate role to your contacts, nothing more.

It can't help you read your WeChat Moments or order takeout.

You could even say that anything you ask him to do has no direct relation to WeChat itself, except that the communication between you and him takes place on WeChat.

Yes, it was attached with transparent tape.

Let's look at something else—

According to a report by The Information in March of this year, WeChat has actually been secretly advancing its own AI Agent project since 2025.

This project has much bigger ambitions—it aims to directly connect with a massive number of mini-programs within the WeChat ecosystem, enabling services such as ride-hailing, food delivery, grocery shopping, and ticket booking (somewhat similar to what Qianwen does). According to reports, it plans to begin gray-scale testing in mid-2026.

In other words, what WeChat is trying to do is completely different from integrating OpenClaw.

But these two things are not contradictory.

OpenClaw addresses the immediate needs of users who already raise shrimp—they need a place to talk to their shrimp.

If WeChat doesn't do it, they'll go to Lark, Telegram, and WeChat Work, causing WeChat to lose user engagement time.

However, the real show will begin once WeChat's own agent is launched. Only when an AI agent can directly manage the capabilities of these mini-programs will WeChat's AI be truly at its full potential.

So this pure tool integration is just the beginning. WeChat is getting you used to talking to AI in your contacts first. Once you're used to it, it will gradually integrate other modules. At that point, user acceptance will be much higher.

First, make it seem normal to you, then make it seem easy to use.

Brother Long has always been like this.

9. We need to gradually get used to the increasing number of non-human contacts appearing in our WeChat contacts.

First there was Yuanbao, then WeChat customer service messages, and now OpenClaw, followed by WeChat's own Agent.

Think about it carefully: before this, every contact in your WeChat address book was a real person. (Enterprise accounts and official accounts don't count; they weren't in the address book).

Now, WeChat officially recognizes a new existence: you can have a non-human contact in your address book.

This shift is quite subtle.

WeChat's core narrative has always been about connecting people, but from the moment it connected with OpenClaw, it has also been connecting people with AI.

This change may have a more profound impact than just functional changes.

Because once users accept that AI can exist in their contacts, what's next?

It's only natural that AI can do some things in WeChat, and that's how things should naturally fall into place.

This change may come sooner than we imagine.

10. WeChat's decision not to develop its own shrimp product was a significant product decision in itself.

Have you ever thought about this: with the resources and technical strength of the WeChat business unit, WeChat could easily create its own shrimp—just like Qclaw, built into WeChat and ready to use out of the box.

But WeChat did not do that.

The decision not to do this is more worthy of analysis than what was done.

If you're just a shrimp, you're just an ordinary shrimp in a battle of a thousand shrimps; but if you're a connector, you become part of the infrastructure.

WeChat does have its own AI Agent project, as mentioned in point 6. In a sense, it's a shrimp with WeChat DNA, a shrimp with unlimited communication bandwidth with WeChat.

So here's the question: when WeChat's own Agent is officially launched, what will be the relationship between it and OpenClaw?

Once WeChat's own agent is launched, its entry point, promotional resources, and position within the WeChat interface will likely be much higher priority than OpenClaw's shrimp.

After all, he's his own son.

This time, WeChat has chosen to support all apps compatible with the OpenClaw protocol, instead of just supporting Tencent's own, making it a neutral platform.

How other shrimp vendors can differentiate themselves from WeChat's own "True Destiny Shrimp" is a question worth considering.

Conclusion

It is easy for us to overestimate the short-term impact of something and underestimate its long-term impact.

WeChat's integration with OpenClaw will not change anything in the short term.

But looking back ten years later, this may be the starting point for WeChat to move from connecting humans and fish to connecting humans with AI.

Let's recall the history of the widespread adoption of QR codes.

When WeChat added the scan function in 2012, the vast majority of Chinese people had no idea what a QR code was. WeChat did not educate users; it simply put the scan function into WeChat.

Then users will wonder: What can this thing do?

Then merchants started posting QR codes, and then mobile payment came along.

Putting it in WeChat is important in itself.

Every step WeChat takes may seem small, but in retrospect, each one is quite precise.

This time too.

However, Brother Long has never been in a hurry.