Atlassian adds visual AI and agents from Lovable, Replit, and Gamma to Confluence

Atlassian has expanded Confluence's capabilities by adding built-in visual AI tools and support for third-party agents from Lovable, Replit, and Gamma. Previously, teams working in Confluence were forced to switch between multiple services to create diagrams, presentations, or prototypes. Now visual materials can be generated directly within the platform — without exporting and importing files.

The new agents from three partners solve different tasks. Lovable specializes in rapid creation of user interfaces and web applications. Replit is a platform for developing and running code directly in the browser.

Gamma generates visual presentations and documents based on text descriptions. All three agents are integrated into Confluence's working context: they can see page content and act based on it. Atlassian has long been moving toward an AI platform for team collaboration.

In 2025, the company introduced Rovo — a set of AI tools on top of its products. The integration of external agents continues this strategy: instead of building all capabilities in-house, Atlassian is opening Confluence as an execution environment for specialized AI tools. For users, this means reducing context switching.

Instead of the chain "formulated in Confluence → opened Replit → wrote code → pasted the result back," the entire cycle can potentially close within a single interface. While the scope of integration is currently limited to three partners, the architecture of open agents itself points to the direction: Confluence is transforming from simply a wiki platform into a working environment where specialized AI executors are connected. How deeply agents will be able to act autonomously will be shown by practical use.