The One Capability the JetBrains Plugin Doesn't Have
Both the VS Code extension and the JetBrains plugin run a local, hidden MCP server for IDE integration, exposing a diagnostics tool. VS Code additionally exposes a tool for running Python in an active Jupyter kernel.
What does the JetBrains plugin expose in that same slot?
The correct answer is "Nothing — JetBrains exposes no code-execution tool at all."
Both IDE plugins run a local, hidden ide MCP server exposing
mcp__ide__getDiagnostics. VS Code additionally exposes
mcp__ide__executeCode for running Python in an active Jupyter
kernel, gated behind an explicit confirmation each time. The
JetBrains plugin exposes no code-execution tool to the model at
all — a concrete, documented capability gap between the two IDE
integrations, not a setting you can turn on.
The distractors invent capabilities the JetBrains plugin doesn't
have: there's no read-only shell tool standing in its place, no
identical executeCode tool hidden behind a trust setting, and no
separate JVM-language execution tool.
Share this question