NumPy array metadata (shape, dtype, ndim)

Type: Feature Request

Hello VS Code team,

Feature request: please add an option to show NumPy array metadata (shape, dtype, ndim) directly in the Variables/Locals view when debugging Python. Current workarounds (Data Viewer, Watch, Debug Console) are helpful but a compact shape/dtype column would greatly speed up debugging numeric code.

Use case:

• Quickly inspect array shapes/dtypes (e.g., frame.shape, frame.dtype) while paused at breakpoints without opening the Data Viewer or adding watches.
• Helpful for image-processing / ML workflows where many arrays are inspected frequently.

Suggested design:

• Add optional compact badges or columns in Variables: name | type | shape | dtype | value-preview.
• Or provide a toggle Python: Show Array Shape/DType in Variables.

Repro steps:

1. Open a Python file using NumPy arrays.
2. Start debugger and pause on a line where arrays exist.
3. Observe Variables view — shape/dtype not visible by default.

Environment:

• VS Code version: 1.109.4
• Python extension: 2026.2.0
• OS: linux

Thanks for considering.

Regards,
Uri ZACKHEM urizackhem@gmail.com

VS Code version: Code 1.109.4 (c3a26841a84f20dfe0850d0a5a9bd01da4f003ea, 2026-02-16T15:35:57.932Z)
OS version: Linux x64 6.17.0-14-generic snap
Modes:

Start by locating the Variables/Locals rendering entry point in the TypeScript debugger extension and reviewing how paused Python values are represented. Determine where NumPy shape, dtype, and ndim could be exposed and how an optional display should behave; done means the requested metadata is visible while paused without requiring the Data Viewer or watches.