Trae is launching some exciting updates:

• We've unified Chat and Builder into one.

• We are introducing @Agent. With @Agent, Trae seamlessly combines multi-agent capabilities and adds MCP integrations to establish a universal communication framework. You can not only build with Builder—our built-in agent—but also create your own custom agents.

• We've expanded our context capabilities (#Web, #Doc, Rules) to provide you with a robust knowledge foundation for all your projects. We categorize rules as a form of context because, similar to web search results and documents, rules provide contextual information for models to understand.

@Agent: Reimaging the Future of Programming

The introduction of @Agent transforms Trae into a versatile team of AI specialists ready at your command. Each Agent possesses unique capabilities designed for specific development tasks, allowing you to call upon the right Agent for their specific task. You can use Trae's built-in agents while also having the freedom to create your own custom agents.

Customize your own agents with Rules

We're introducing Rules with @Agent. You can now define Rules that shape your agent behaviors—from coding style preferences to complexity limits and architectural guidelines. This empowers developers to create custom agents tailored to their unique workflows and ensures Human-AI collaboration aligns perfectly with your established practices and standards.

MCP Integration on Trae

This is just the beginning of @Agent's potential. Beyond these individual capabilities, we're integrating MCP (Model Context Protocol) into Trae, which establishes a universal communication framework that enables seamless interactions between Trae's native Agents and third-party extensions. MCP intelligently coordinates the optimal combination of agents based on your specific requirements, orchestrating them into a unified problem-solving team.

Combining Chat and Builder: Chat is the Next Generation Terminal

We're unifying Chat and Builder in this new version, bringing our "Chat is the Next Generation Terminal" vision to life. We believe natural language communication will become the primary interface for human-AI collaboration, superseding traditional command-line interfaces.

In the previous version, switching between Chat (conversational programming) and Builder (AI-driven project development) often results in context loss and redundant explanations. The new experience preserves your complete context as you move between conversational programming and project building. Builder, the built-in agent, retains awareness of all requirements, discussions, and code snippets - no more repetitive explanations or lost context.

This integration represents more than a feature update. It's a fundamental reimagining of how developers and AI can collaborate more efficiently in a modern development environment.

@Agent and #Context: the Design Philosophy Behind Our Symbols

This functional upgrade embodies an interesting design philosophy. We've frequently heard from our developer community: "Why does Context use # instead of @?" The answer reveals our deeper approach to AI-human collaboration.

From day one, Trae has aimed to be a collaborative AI partner, not merely an assistant. Our choice of symbols goes beyond mere syntax—it's about bringing familiar social interactions into the development environment. These carefully considered design choices reflect our commitment to creating an AI coding environment that feels both intuitive and powerful for modern developers. When you @Agent in Trae, you're not just executing a command; you're engaging a specialized collaborator, mirroring how you'd mention a teammate on social platforms.

For consideration: Similarly, #Context unifies related elements into a cohesive environment, adopting the familiar hashtag concept from content organization systems. This release expands #context capabilities with specialized variants like:

• #Web - Seamlessly incorporate internet resources and external references

• #Doc - Integrate project documentation, specifications, and technical resources

This structured approach enables Trae to understand your development intentions with unprecedented precision, creating a knowledge foundation that minimizes misunderstandings and accelerates your workflow. By categorizing different types of context, you gain finer control over how information shapes your AI collaboration experience.

This release transforms Trae from a reactive coding assistant to a proactive development partner—one that thinks alongside you, offers strategic suggestions, and anticipates your next steps. As routine development tasks become increasingly automated, you're free to focus on creative work or even contribute to the Trae Agent ecosystem.

We're building more than a product—we're creating an ecosystem where infinite intelligence emerges through collaboration. Developers and third parties can now contribute freely and collaboratively by creating, sharing and customizing different agents that can expand Trae's capabilities exponentially.

Join us in reshaping the future of development! 🚀