China's AI Giants Battle for Skill Store Dominance

Chinese tech giants Tencent, Alibaba, and ByteDance have launched competing 'Skill' stores within their respective AI agent platforms. This move marks a critical escalation in the race to define how artificial intelligence agents interact with users and tools.

The term Skill refers to structured instruction files that guide AI agents. These files dictate which tools an agent should call, how to handle specific scenarios, and what output standards to follow. Essentially, they act as operation manuals for autonomous software.

• Major Players Involved: Tencent, Alibaba, and ByteDance led the initial rollout in March.
• New Entrants: Zhipu, Meituan, and Xiaohongshu joined the fray shortly after.
• Core Function: Skills enable agents to execute complex, predefined workflows reliably.
• Current Model: Most skills are free, except for ByteDance's Coze platform.
• Strategic Goal: Controlling the distribution channel for AI capabilities.
• Market Shift: Moving from model-centric to application-centric competition.

Defining the Skill Economy

To understand the significance of this development, one must first grasp what a Skill actually is. In the context of Large Language Models (LLMs), a skill is not merely a plugin. It is a comprehensive, structured directive file.

This file contains precise instructions on tool invocation, decision-making logic, and output formatting. When an AI agent reads this file, it follows a preset path to complete tasks. This reduces hallucination and increases reliability in professional settings.

Consider a senior product manager who has perfected the art of writing product requirement documents (PRDs). They can encapsulate their entire workflow into a single Skill. Once uploaded, any user’s AI agent can adopt this Skill. The agent then generates PRDs following the exact same high-quality framework.

This democratizes expertise. It allows specialized knowledge to be packaged and distributed like software applications. The analogy is clear: if LLMs are the operating systems, Skills are the apps that run on them.

Initially, these Skills were shared on developer communities like GitHub or ClawHub. These platforms served as early hubs for uploading, searching, and downloading technical assets. However, the audience was limited to developers and tech enthusiasts.

The entry of major internet companies changes the scale entirely. By integrating Skill stores directly into consumer-facing agent platforms, these companies aim to reach millions of non-technical users. This shifts the paradigm from code sharing to capability sharing.

The Strategic Race for Distribution Control

Why are these tech behemoths investing heavily in free repositories? The answer lies in traffic entry points. In the digital economy, whoever controls the distribution channel controls the user.

By March of this year, Tencent, Alibaba, and ByteDance had all launched their own Skill stores. This was not a coincidence but a coordinated strategic pivot. Following their lead, Zhipu, Meituan, and Xiaohongshu entered the market within two months.

This rapid expansion highlights the urgency of the situation. Internet giants, large model companies, local life service providers, and content platforms are all vying for the same position. They are fighting to become the default interface for AI interactions.

• Tencent: Leveraging its vast social media and gaming ecosystem.
• Alibaba: Integrating Skills into enterprise and e-commerce workflows.
• ByteDance: Using its recommendation algorithms to surface popular Skills.
• Meituan: Focusing on local service automation and delivery logistics.
• Xiaohongshu: Targeting lifestyle and content creation utilities.

The underlying logic is simple. If users rely on a specific platform to find and install Skills, that platform becomes indispensable. It creates a lock-in effect similar to app stores in the mobile era.

Currently, most platforms offer Skills for free. ByteDance’s Coze platform has begun testing paid Skills, but others remain open. This suggests that immediate monetization is not the primary goal.

Instead, the focus is on ecosystem growth. A rich library of Skills makes an agent platform more attractive. More users attract more developers, creating a virtuous cycle. This network effect is crucial for long-term dominance.

Implications for Developers and Businesses

For developers, this trend presents both opportunities and challenges. On one hand, there is a new marketplace for monetizing expertise. On the other, fragmentation may occur if each platform uses proprietary formats.

Businesses should monitor these developments closely. Adopting a multi-platform strategy might be necessary to ensure broad reach. Alternatively, partnering with a dominant player could provide faster access to users.

The rise of Skills also lowers the barrier to entry for AI adoption. Companies no longer need to build custom models from scratch. They can assemble existing Skills to create tailored solutions quickly.

This shift empowers non-technical staff. Marketing teams, HR departments, and customer support units can customize their AI assistants without coding. This decentralization of AI development is a significant cultural shift.

However, quality control remains a concern. With thousands of Skills available, how do users verify reliability? Platform providers will need robust rating and review systems. Security audits will also be essential to prevent malicious code injection.

Looking Ahead: The Future of Agent Interaction

The competition among Chinese tech giants mirrors earlier battles in the smartphone and cloud computing sectors. History suggests that consolidation will eventually occur, but the interim period will be marked by intense innovation.

We can expect to see greater standardization efforts. Industry groups may push for universal Skill formats to ensure interoperability. This would benefit developers by allowing them to write once and deploy everywhere.

Furthermore, the integration of Skills with real-world actions will deepen. Agents will not just generate text; they will execute transactions, book appointments, and manage supply chains. The stakes for reliability and security will increase accordingly.

For Western audiences, this serves as a preview of global trends. Similar dynamics are emerging in the US and Europe, albeit with different players. Understanding the Chinese model provides valuable insights into the future of AI distribution.

The battle for the Skill store is not just about technology. It is about defining the economic structure of the AI era. Whoever wins this battle will shape how humans and machines collaborate for years to come.

Gogo's Take

• 🔥 Why This Matters: This represents the transition from AI as a novelty to AI as infrastructure. By controlling the 'app store' for agents, these giants are positioning themselves as the gatekeepers of productivity. For businesses, this means AI customization will become as easy as installing an app, drastically reducing implementation time and cost.
• ⚠️ Limitations & Risks: Fragmentation is the biggest risk. If Tencent, Alibaba, and ByteDance use incompatible Skill formats, developers face a nightmare of maintaining multiple versions. Additionally, the lack of immediate monetization for most platforms raises questions about sustainability. Who pays for the compute costs of running these complex Skills?
• 💡 Actionable Advice: Developers should prioritize building modular, platform-agnostic Skills where possible. Keep your core logic separate from platform-specific wrappers. For businesses, start experimenting with agent-based workflows now using free tiers on platforms like Coze or Tongyi Qianwen to understand the potential before committing to a specific ecosystem.