Atlassian Corporation, known for its collaboration tools such as Jira, Loom and Confluence, has introduced new AI-driven features in its Confluence knowledge management platform. The updates are designed to support the transformation of content into visual formats while retaining context and reducing the need for manual reformatting.
A new feature, “Remix”, currently in open beta, allows users to convert Confluence pages into visual outputs such as charts, infographics and presentation-ready summaries. Users can select content including paragraphs, tables or full documents and generate inline visuals. The feature also provides format suggestions based on content type and usage patterns within the Teamwork Graph.
Remix is designed to maintain a single source of truth by layering visuals on top of linked original content, without requiring exports or the creation of additional pages. At launch, it supports formats including data visualisations, infographics, scorecards and charts, with further formats planned.
Atlassian has also introduced third-party agents built on the Model Context Protocol (MCP), connecting Confluence with partner tools. The initial set of agents, launched on 13 April, includes:
- The Lovable Agent, which converts product specifications into UI prototypes
- The Replit Agent, which turns technical documents into starter applications
- The Gamma Agent, which transforms text such as meeting notes or status pages into presentations
These agents can be launched from Confluence, where they access page content and metadata and pass context into partner tools. The resulting outputs link back to the original Confluence page. The system is powered by Rovo and built on MCP, enabling Confluence content to be used across integrated tools while retaining context.
Atlassian states that setup is designed to be simple, with administrators able to enable partner MCP servers through Atlassian Administration, allowing agents to appear in the Rovo directory. The company plans to expand the range of partner integrations over time, further extending its AI-enabled ecosystem.
