Atlassian introduced the Agentic Multiplayer Protocol (AMP) at the Team ‘26 Europe event. This platform update aims to integrate AI agents into corporate workflows as active team members, allowing them to collaborate with humans within a governed digital environment.
The company launched Rovo Work, a new mode for its AI assistant designed for complex, multi-step tasks. Unlike standard models, Rovo Work can research and learn new skills, such as video synthesis, if the required information is missing from its training data.
Atlassian is also integrating Loom to allow users to visually instruct AI agents by recording screen actions and verbal commands. This feature translates visual cues, such as clicking buttons or highlighting interface elements, into prompts for the agents.
The platform now includes Rovo Code Search and expanded Data Context, which connect the Teamwork Graph to external platforms like Databricks, Snowflake, and BigQuery. This provides agents with a comprehensive view of both software development and business data.
Atlassian’s Model Context Protocol (MCP) server now supports 200 tools and processes approximately 15 million tool calls daily. The system allows agents to write information back to the platform, enabling real-time collaboration between humans and AI.
