Xiaohongshu launches the competition for discovery access in the RED Skill AI era
```
On June 8, Xiaohongshu officially launched the RED Skill feature.
When some creators publish their posts, they can attach AI Skill components below the content. Users can click to copy the command and install it in Agent products that support the related capability. The platform simultaneously released official support initiatives and a curated Skill ranking list.
Viewed on a broader timeline, this is not an isolated product update.
Over a month ago, on April 30, Xiaohongshu issued a company-wide internal letter, announcing the establishment of a first-level department called Dots, centered on AI.
This department does not belong to any particular business line, but reports directly to the new president Conan, and has its own independent budget and personnel authority, standing alongside core businesses such as community, e-commerce, and commercialization. At the same time, Xiaohongshu established an Enterprise Intelligence Department, preparing for the AI era across organizational, technical, and efficiency dimensions.
In the past few years, Xiaohongshu’s actions on AI were relatively restrained, mainly focused on content production, search, and recommendation. But with the formation of Dots, AI began to rise from a tool capability to a company-level strategic direction.
RED Skill has become a specific implementation of this shift on the community side.
The problem it addresses is straightforward: lowering the threshold for AI Skill discovery and dissemination.
Over the past half year, Skill has gradually become one of the most active topics in the Agent field.
Simply put, a Skill is a structured operational guideline for AI Agents. When an Agent reads this instruction, it can follow the preset workflow to complete specific tasks.
For example, an experienced product manager can package their method for writing requirement documents into a Skill, and other users’ Agents, after installing it, can also complete work by following a similar process.
As the number of Skills keeps growing, how users discover them has become a new issue.
The earliest players in this role were mainly developer communities like GitHub, ClawHub, etc. Entering 2026, more and more Agent platforms began building their own Skill ecosystems and capability markets. Tencent, Alibaba, ByteDance, and others have all launched relevant capability entry points for developers and users, while Zhipu, Meituan, and others are also keeping up.
Compared to these platforms, Xiaohongshu’s entry point is quite different.
Most Agent platforms aim to build Skill operating ecosystems, while Xiaohongshu is closer to Skill discovery and dissemination scenarios from the perspective of a content community.
Understanding this point may make it easier to grasp why RED Skill emerged.
As Agents gradually become the new entry point for information processing, new traffic scenarios are forming around Skill discovery, dissemination, and installation chains. Xiaohongshu holds a relatively unique asset—a continuously growing technical content community.
This did not happen overnight.
According to official data, in the past year, Xiaohongshu’s tech content publication volume grew over 100% year-on-year, and the number of creators increased over 200% year-on-year.
Content related to “Build in Public” within the platform has accumulated over 1.1 million posts, with those born after 2000 and 2005 being the main participants.
Meanwhile, the number of active developers on the platform has exceeded 160,000, with more than 90% having developed more than one product in a year.
Regarding Skills, Xiaohongshu revealed that there are already 300,000 creators of AI Skill-related content on the platform, with related topics having received over 600 million exposures. After RED Skill went online, nearly 1,000 original Skills have been published within the community.
Behind these numbers, the change in demographic structure is even more noteworthy.
This April, Xiaohongshu held its first Hackathon Summit, with 200 shortlisted developers engaging in a 48-hour closed development. Over 60% were born after 2000, and the youngest participant was only 13 years old.
Compared to the previous generation of developers, these AI natives are more accustomed to leveraging AI to complete development work, and are more willing to openly share their product-building process.
Project progress, product inspiration, user feedback, and even failures all become part of public discussion.
This “Build in Public” culture is becoming a key feature of Xiaohongshu’s tech community.
Xiaohongshu’s science and technology vertical leader Sanbing once summarized it as: “Let creators show their progress, entrepreneurs find partners, investors discover projects.”
From this perspective, RED Skill is more like adding a standardized dissemination pathway for Skills atop the existing content ecosystem.
Of course, RED Skill currently still has obvious limitations.
After users discover Skills on Xiaohongshu, subsequent installation and operation still requires jumping to external Agent products. Xiaohongshu cannot directly sense whether users have completed installation, whether they continue to use, or how effective actual operations are.
Another detail: the “Number of Users” displayed on the details page currently counts those who clicked the “Go Use” button, not those who actually finished installing or invoked the Skill.
This means there is still significant conversion loss between discovery and actual usage.
At this stage, RED Skill acts more as a Skill discovery entry, rather than a full Skill ecosystem platform.
But as the Skill formats gradually unify and compatibility among different Agents strengthens, the ways Skills are disseminated may still change.
The role of the content community—whether as a traffic entry point, product marketplace, or developer community—is not yet clear.
However, what is certain is that as the Agent ecosystem develops, the discovery and dissemination of Skills is becoming a new competitive direction.
RED Skill is just the beginning of this transformation. Who will truly master users in the future depends on whether a complete closed loop can be formed in steps like discovery, installation, operation, and retention.
And Xiaohongshu is trying to claim a part of that position.
Risk Warning and DisclaimerThe market involves risks; investment should be cautious. This article does not constitute personal investment advice and does not take into account the specific investment goals, financial conditions, or needs of individual users. Users should consider whether any opinions, views, or conclusions herein fit their specific circumstances. Any investments made accordingly are at one's own risk. ```