Atlassian

    Facilitator notes

    The tools land best when people try them on real work, not demos. The goal isn't to master every tool — it's to build the instinct for matching the right tool to the task in front of you.

    Before you start

    • Sort out accounts, access, and setup before the session — nothing kills momentum like tool troubleshooting.
    • Ask each person to bring one real task they can actually build during the session.

    Tools

    After learning tooling basics, help teams apply AI to an actual delivery process. Map your sessions to the way your own teams discover, plan, build, measure, and operate using your own tools. What you'll see in the following sections is Atlassian's own curriculum, based on the tools we use. You can customize this for your own environment.

    Tooling philosophy

    Before you teach tools, explain how people should choose between them. At Atlassian, we use three rules:

    Use hands-on practice

    Run live walkthroughs and labs where people use the tools on realistic tasks.

    Match tool to task

    No single tool is right for every task. Teach people when to switch tools.

    Treat tokens like a budget, not a buffet

    Teach people to manage context, prompts, and cost instead of treating tokens as unlimited.

    Our core tooling

    List the tools your teams should use and explain when to use each one. Atlassian organizes tools this way:

    1

    Fast ideation and customer testing

    Replit and Figma are our go-to tools for moving fast in the early stages — spinning up working prototypes from a prompt, iterating on designs with AI, and getting real feedback before committing to a full build.

    2

    Technical prototyping

    Our Prototyping Sandbox repo is where ideas graduate into engineering-ready code. Use it for more complex builds that need structure, version control, and a clear path to handoff.

    3

    Production-grade development and automation

    Rovo CLI, Claude, Codex, and Cursor are examples of tools we use to query work data, edit real codebases with AI, automate with natural language, and reason through hard problems at speed.

    Production tooling

    ToolWhat it isBest for
    CursorAI-first code editor with a strong chat and inline edit flowQuick iteration on a single file or screen, fast hands-on prototyping
    CodexOpenAI agent for multi-step coding workComplex pipelines, multi-step tasks, longer-running agentic work
    Rovo DevRovo CLIAtlassian-built coding agent with Teamwork Graph integrationADS-accurate prototypes, Confluence/Jira-connected work, internal data
    Rovo StudioApp BuilderNo-/low-code way to build Forge apps inside Rovo StudioTurning ideas into working Forge apps for Jira, Confluence, and more without full-time dev skills
    Claude CodeAnthropic coding tool — strong on reasoning, careful with changesLarger refactors, careful multi-file edits, deliberate agent work

    Want to go deeper?

    Use these Atlassian examples if you want more detail:

    Developer EnvironmentsRovo CLIApp Builder