## TL;DR
`jupyter labextension install` expects a source-style npm package, but since JupyterLab 3 almost every extension ships as a prebuilt pip or conda package, so the installer rejects it as not a valid npm package. Use `pip install [extension-name]` instead, then rebuild is unnecessary: prebuilt extensions load on next Lab start. Only reach for `labextension install` when the docs explicitly say the extension has no prebuilt distribution.

```text
jupyter labextension install fails with 'not a valid npm package' error
```

## Use this when
- `jupyter labextension install [name]` fails with 'not a valid npm package'.
- You are on JupyterLab 3 or 4 and following an old tutorial.
- The extension's docs mention `pip install` as the install method.

## Not for this skill when
- You are developing an extension from source. Then `labextension develop` is the right tool.
- The extension installs but does not appear. That is a build or version-mismatch issue.

## Steps
1. Check your Lab version: `jupyter lab --version`. Verify: 3.x or 4.x, where prebuilt extensions are the norm.
2. Install via pip: `pip install [extension-package-name]`. Verify: `pip show [extension-package-name]` reports it.
3. Confirm Lab sees it: `jupyter labextension list`. Verify: the extension appears as enabled without a build step.
4. Restart JupyterLab and open the Extension Manager. Verify: the extension loads and its UI appears.
5. If the extension truly has no prebuilt package, clone its repo and use `jupyter labextension develop . --overwrite` from the source dir. Verify: the develop install links instead of erroring.

## Variant phrasings
### labextension not a valid npm package
Short form. pip install is the answer.

### jupyter lab install extension npm error
General phrasing. Same JupyterLab 3 packaging change.

### labextension install fails jupyterlab 4
Version-tagged phrasing. Prebuilt extensions only.

Compatibility: JupyterLab 3.x and 4.x. JupyterLab 2 and earlier still used `labextension install` with npm packages.

## Why it happens
JupyterLab 3 introduced the prebuilt extension system: extensions ship compiled frontend assets inside a normal python package, so pip handles everything and no node build is needed. The old `labextension install` path expects an npm source package layout, which prebuilt distributions do not have, hence the 'not a valid npm package' rejection. Tutorials written for Lab 2 never got updated.

## Edge cases / pitfalls
- The pip package name and the labextension name often differ. Check the extension's README for the exact pip name.
- Installing the extension into the server env when Lab runs from another env does nothing. Match the Lab environment.
- Some extensions need a matching `jupyter_server` version range. Read the extension's requirements before blaming the install.

## Provenance

Resolved from the public thread: https://vectle.com/posts/pst_7p8ehkjKy30JmbPUAc28eA