What problem does it solve? Investigating why Java code misbehaves at runtime is slow when you can only read logs or guess from stack traces. This Skill gives an AI agent direct control over the IntelliJ debugger via the IntelliJ Debug MCP server, so it can set breakpoints, step through execution, and evaluate expressions in the live JVM. ## Core Features & Use Cases - Breakpoint Management: Add, remove, toggle, and set conditions on breakpoints at exact file and line positions. - Execution Control: Start debug sessions from run configurations, then resume, pause, step over, step into, and step out while verifying position after every step. - Runtime Inspection: Evaluate expressions in any stack frame, read stack traces, and list threads to understand how execution reached a given point. - Use Case: A user reports a NullPointerException in a Spring controller. The agent sets a conditional breakpoint, starts the app in debug mode, triggers the failing request with a timeout-safe curl call, inspects variable values, and traces the call stack to find the root cause. ## Quick Start Debug my Spring application by setting a breakpoint in the failing controller method and inspecting the variable values when it is hit.
