Junie is a coding agent officially released by JetBrains on June 17, 2026. Designed to go beyond simple code suggestion and auto-completion, it enables developers to continuously plan, implement, test, debug, and review software development tasks. It can be used not only in JetBrains IDEs but also in CLI, GitHub Actions, and GitLab environments. Its key feature is the ability for developers to select models tailored to their specific tasks. Just as text generation models read the context around a sentence to generate the next content, Junie reads the semantic index of code built by the IDE, along with project structure, test tools, and execution information from the debugger, to determine which code needs to be changed. While existing coding assistance tools are useful for quickly writing short functions or repetitive code blocks, developers often have to re-provide full context when interpreting requirements spanning multiple files or tracking defects that only appear during execution. Junie’s Advanced Plan mode addresses this by documenting requirements, technical designs, and step-by-step tasks before implementation begins, thereby reducing the risk of losing focus during long-duration tasks. This approach is akin to creating blueprints and process schedules before construction, rather than generating code immediately. By reviewing the plan, developers can adjust the implementation scope and approach, allowing them to manage the risk of the agent making large-scale changes based on incorrect assumptions. Agentic debugging is a distinct differentiator for Junie. Rather than merely reading error logs and inferring fixes, it actively manipulates the IDE debugger to set breakpoints (Breakpoint) and inspect stack frames (Stack frame) and runtime values. This makes it well-suited for issues that require evidence from execution time, such as state transition errors where the cause is not obvious from source code alone, incorrect conditional branches, or variable values that differ from expectations. By combining IDE semantic code navigation with test execution, Junie enables a unified workflow that links root cause exploration, code modification, and regression test verification. Bioinformatics research teams can utilize Junie for maintaining analysis pipelines and verifying reproducibility. For example, when handling exceptions that occur only with specific inputs in Python-based preprocessing code, developers can first document the reproduction conditions and fix steps using Advanced Plan, then verify the fix by checking runtime values via the JetBrains IDE debugger and executing tests. Additionally, after writing data processing modules in the CLI, connecting Junie to automated testing processes in GitHub Actions or GitLab allows for extending change reviews and failure analysis beyond the local development environment. However, supported IDEs, available models, data handling policies, permission scopes for automation environments, and detailed installation procedures should be verified through official documentation.

Junie is a coding agent officially released by JetBrains on June 17, 2026. Designed to go beyond simple code suggestion via autocomplete, it continuously handles planning, implementation, testing, debugging, and review of software development tasks. It can be used not only in JetBrains IDEs but also in CLI, GitHub Actions, and GitLab environments. Its key feature is that developers can select a model tailored to their specific task. Just as text generation models read the context around a sentence to generate the next content, Junie leverages the semantic index of code built by the IDE (Sem