Atlassian introduces AI visual tools and partner agents to Confluence, just a month after laying off 1,600 employees.

      In summary: Atlassian is launching Remix, a visual AI tool in open beta that converts Confluence pages into charts, infographics, and scorecards without needing users to open additional applications. This rollout also includes three partner agents based on the Model Context Protocol, which will transfer Confluence content directly into Lovable, Replit, and Gamma starting April 13. This announcement follows Atlassian's recent layoff of 1,600 employees to fund AI initiatives.

      Knowledge management software faces challenges in presentation. Teams invest significant time documenting choices, specifications, and meeting results in Confluence, only to spend a similar amount of time manually reformatting that content into charts, prototypes, and presentations needed for different audiences. Atlassian aims to bridge this gap with two interconnected announcements this Wednesday: a visual generation tool called Remix that links output to its source, and a set of ready-made agents that transfer Confluence content directly to partner applications.

      Remix: self-transforming documentation

      Remix, now available in open beta, enables teams to select any content from a Confluence page—from a paragraph to a table or a full document—and direct the tool to create a visual representation from it. At launch, it supports output formats such as data visualizations, infographics, scorecards, and charts, with plans to expand the formats over time. The generated visual is displayed alongside the original content and linked to it, so it updates automatically as changes occur to the underlying page, eliminating the need for separate exports or file management processes.

      The intelligence behind Remix’s format suggestions comes from the Teamwork Graph, Atlassian’s unified data layer that utilizes over 100 billion data points from Jira, Confluence, and associated enterprise tools. Instead of requiring users to choose a format manually, Remix employs this graph to suggest the most suitable visual type based on the content's structure and the organization’s usage patterns; for example, a quarterly roadmap page could lead to a scorecard, while a dataset might result in a chart.

      Sanchan Saxena, Atlassian’s senior vice president of product for the Teamwork Collection, described the tool as an effort to make the platform more seamless: “With Remix and agents in Confluence, a single page becomes the foundation for whatever comes next: a clear narrative for leaders, a prototype for builders, or a guide for customers, all derived from the same source of truth.”

      Three agents that bridge application boundaries

      While Remix keeps outputs within Confluence, the partner agents introduced alongside it are intended to transfer content from Confluence into specialized tools without any manual copying or custom integration. Three agents will launch on April 13: Lovable, which turns a product specification into a functional user interface prototype; Replit, which transforms a technical document into a starter application that engineers can fork and develop; and Gamma, which converts meeting notes or status updates into a polished presentation.

      Each agent is activated directly from a Confluence page via Rovo Chat. When initiated, the agent reads the page’s content and metadata, including authorship, project affiliation, and decision context, and transfers all of this information into the partner tool without requiring the user to reconstruct that context manually. The produced artifact—be it a prototype in Lovable, a codebase in Replit, or a deck in Gamma—remains linked to the original page, preserving the connection between documentation and output.

      For administrators, setup necessitates no custom scripting. Enabling a partner’s Model Context Protocol server in Atlassian Administration takes a few minutes, after which the agent appears in the team's Rovo directory, pre-configured by the partner and inheriting existing permissions and contextual settings.

      MCP as the open standard

      The technical foundation for the partner agents is the Model Context Protocol, an open standard that has quickly become the binding element of the agentic software ecosystem. Atlassian's decision to utilize MCP instead of a proprietary integration layer sends a strategic message: any partner can create an agent that integrates with Confluence content without waiting for Atlassian to establish a customized connection. The protocol is open and the server details are documented, making the entry barrier into this ecosystem one of technical capability rather than a commercial agreement with Atlassian.

      The three launch partners are intentionally designed to cover different use cases. Replit, which is also featured as a launch partner in Anthropic’s newly unveiled enterprise software marketplace, represents the developer workflow; Lovable caters to product design and prototyping processes; while Gamma addresses executive communication needs. Together, they encompass the three primary audiences for which Confluence documentation typically needs to be reformatted to become actionable.

      The AI pivot in context

      In March, Atlassian laid off 1,600 employees, roughly 10% of its global workforce, with CEO Mike Cannon-Brookes indicating that the savings would be redirected towards AI investments and enterprise sales. The company also replaced its chief technology officer, splitting the role between two executives: Taroon Mand