According to a recent LinkedIn post from Manufact (YC S25), the company is highlighting new functionality that links its mcp-use framework with CopilotKit to render MCP Apps inside a product’s chat interface. The post suggests this integration allows AI agents to respond with interactive elements such as maps or charts instead of text, while avoiding custom frontend renderers.
The company’s LinkedIn post describes a three-step setup in which developers define a tool on an MCP server with an attached React view, register the server URL in the CopilotKit runtime, and then invoke the tool from the agent so CopilotKit renders the view. The post also notes that because the client app only needs the server URL, updated views can be deployed without redeploying the frontend, potentially streamlining iteration.
For investors, this content implies a focus on improving developer experience and speed of deployment in AI-enabled applications, areas where ease of integration can influence adoption. If this workflow reduces implementation friction for customers building chat-based interfaces, it could strengthen Manufact’s positioning in the tooling layer around agentic AI systems.
The post further references a detailed tutorial authored by the company’s founding engineer, indicating an effort to provide technical depth and documentation for early adopters. This emphasis on developer education may help create an ecosystem around the tooling, which in turn could support longer-term customer stickiness and potential expansion opportunities if the approach gains traction in production environments.