Junie /demo is a new mode in Junie CLI. Describe what you want to check, and Junie builds and launches your app, interacts with its UI, and records what happens.
The intended benefit is to shift routine, click-based checks away from the developer, while keeping the final result open for inspection.
WHAT YOU GET FROM A DEMO RUN
The run produces an HTML report, screenshots, and a video you can review or share. The HTML report compiles the request, the result, the video, and the screenshots, showing which checks passed, failed, or remained incomplete. TechStaged has also covered Policy window open as Lehane calls for stronger AI safety evidence, shared standards, and durable policy action.
HOW THE DEMO RUNS IN PRACTICE
A demo can be watched live as it navigates the UI, showing the actual clicks and keystrokes performed by the agent. A prompt defines a target or expected outcome to guide the run.
The process also includes explanatory slides and subtitles in the produced video to assist reviewers and teammates.
ADOPTION AND IMPACT IN JETBRAINS WORKFLOWS
JetBrains connected the demo agent to GitHub Actions, using it across multiple internal workflows. The team reports more than 1,500 unique PRs having run Junie /demo and over 2,100 demo videos created.
The approach supports release smoke tests and feature demonstrations by recording outcomes alongside artifacts.
GETTING STARTED AND SETUP
Set up /demo in a repository, review the generated configuration, and start with a small feature. The environment can be Docker-based, with an internal YAML workflow outlining the plan and steps.
A completed configuration may require a JUNIE_API_KEY secret for CI usage. The setup guide suggests selecting changes from a branch, session, working tree, or last commit and then prompting Junie with a specific check.
WHAT QUALIFIES AS A SUCCESSFUL DEMO
A demo run includes both the observed behavior and an explicit verdict in prompts. In the examples, only a PASS outcome is considered a success, while other verdicts can fail the check.
WHAT THE DEMO ENVIRONMENT LOOKS LIKE
The demo environment is a Docker container based on Debian Bookworm. The base image includes Chromium, Node.js, xterm, a virtual desktop via Xvfb, a window manager, and utilities for screenshots, keystrokes, and video capture.
MODEL HANDLING AND AVAILABILITY
Junie selects a model for /demo from an available order: GPT-5.6 SOL, GPT-6 Astra, GPT-5.5, then GPT-5.4. All models run with High reasoning effort in /demo during the run.
RELATED COVERAGE
- Policy window open as Lehane calls for stronger AI safety evidence, shared standards, and durable policy action
- Apple Watch Ultra 4 arrives with Audio Intelligence and enhanced health sensing
- CodeQL 2.27.0 adds Linux ARM64 support and broader language coverage
- Paul Christiano joins OpenAI Foundation Board and its Safety and Security Committee
- Software articles
SOURCES
- The JetBrains Blog: Your diff has a demo now with Junie /demo Published · Primary source
