pnpm crash during initial run

I have zero experience with JavaScript and pnpm so a package manager crash doesn't mean much to me.
It all seemed pretty straight forward so I you have any idea what I'm doing wrong let me know.

Thanks
```
[cloudflare-os] pnpm --version
11.20.0

[cloudflare-os] git log | head
commit e1ab8fbd4f609aff7ede9d490bafe1bcf9b2a682
Merge: bf7f762 2810fe1
Author: Kenton Varda
Date: Wed Aug 5 12:13:51 2026 +0000

Merge branch 'kenton/fix-deploy-link' into 'main'

Fix deploy link in readme.

See merge request cloudflare/developers/gadgets!305

[cloudflare-os] cat /etc/redhat-release
Red Hat Enterprise Linux release 9.8 (Plow)
```

```
[cloudflare-os] pnpm run-local
Scope: all 27 workspace projects
✓ Lockfile passes supply-chain policies (802 entries in 12.1s)
Lockfile is up to date, resolution step is skipped
Packages: +664
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Downloading @phosphor-icons/react@2.1.10: 5.30 MB/5.30 MB, done
Downloading @emoji-mart/data@1.2.1: 5.38 MB/5.38 MB, done
Downloading @oxlint/binding-linux-x64-gnu@1.76.0: 6.71 MB/6.71 MB, done
Downloading monaco-editor@0.52.2: 18.44 MB/18.44 MB, done
Packages are hard linked from the content-addressable store to the virtual store.
Content-addressable store is at: /home/hpcpony/.local/share/pnpm/store/v11
Virtual store is at: node_modules/.pnpm
Downloading @cloudflare/workerd-linux-64@1.20260731.1: 37.34 MB/37.34 MB, done
Downloading @img/sharp-libvips-linux-x64@1.2.4: 7.53 MB/7.53 MB, done
Downloading @img/sharp-libvips-linux-x64@1.3.1: 8.11 MB/8.11 MB, done
Progress: resolved 664, reused 0, downloaded 662, added 664, done
node_modules/.pnpm/core-js-pure@3.49.0/node_modules/core-js-pure: Running postinstall script...
node_modules/.pnpm/protobufjs@7.6.5/node_modules/protobufjs: Running postinstall script...
node_modules/.pnpm/sharp@0.34.5/node_modules/sharp: Running install script...
node_modules/.pnpm/esbuild@0.28.1/node_modules/esbuild: Running postinstall script...
[ELIFECYCLE] Command failed.
node_modules/.pnpm/workerd@1.20260731.1/node_modules/workerd: Running postinstall script...
[ERROR] Command failed with exit code 1: '/home/hpcpony/.local/share/pnpm/store/v11/links/@pnpm/exe/directory/a1b...375/node_modules/@pnpm/exe/pnpm' install

pnpm: Command failed with exit code 1: '/home/hpcpony/.local/share/pnpm/store/v11/links/@pnpm/exe/directory/a1b...375/node_modules/@pnpm/exe/pnpm' install
at getFinalError (file:///home/hpcpony/.local/share/pnpm/store/v11/links/@pnpm/exe/directory/a1b...375/node_modules/@pnpm/exe/dist/pnpm.mjs:89001:14)
at makeError (file:///home/hpcpony/.local/share/pnpm/store/v11/links/@pnpm/exe/directory/a1b...375/node_modules/@pnpm/exe/dist/pnpm.mjs:91308:21)
at getSyncResult (file:///home/hpcpony/.local/share/pnpm/store/v11/links/@pnpm/exe/directory/a1b...375/node_modules/@pnpm/exe/dist/pnpm.mjs:93152:10)
at spawnSubprocessSync (file:///home/hpcpony/.local/share/pnpm/store/v11/links/@pnpm/exe/directory/a1b...375/node_modules/@pnpm/exe/dist/pnpm.mjs:93112:14)
at execaCoreSync (file:///home/hpcpony/.local/share/pnpm/store/v11/links/@pnpm/exe/directory/a1b...375/node_modules/@pnpm/exe/dist/pnpm.mjs:93042:23)
at callBoundExeca (file:///home/hpcpony/.local/share/pnpm/store/v11/links/@pnpm/exe/directory/a1b...375/node_modules/@pnpm/exe/dist/pnpm.mjs:95570:23)
at boundExeca (file:///home/hpcpony/.local/share/pnpm/store/v11/links/@pnpm/exe/directory/a1b...375/node_modules/@pnpm/exe/dist/pnpm.mjs:95547:49)
at sync2 (file:///home/hpcpony/.local/share/pnpm/store/v11/links/@pnpm/exe/directory/a1b...375/node_modules/@pnpm/exe/dist/pnpm.mjs:95706:10)
at runPnpmCli (file:///home/hpcpony/.local/share/pnpm/store/v11/links/@pnpm/exe/directory/a1b...375/node_modules/@pnpm/exe/dist/pnpm.mjs:252534:5)
at runDepsStatusCheck
```(file:///home/hpcpony/.local/share/pnpm/store/v11/links/@pnpm/exe/directory/a1b...375/node_modules/@pnpm/exe/dist/pnpm.mjs:254302:7)

Start by reproducing the initial run with `pnpm run-local` on the reported Red Hat Enterprise Linux 9.8 environment and inspect the full output for the `workerd` postinstall failure. The issue does not identify a source file or test, so first trace the `run-local` entry point and the `workerd` installation step. Done means explaining or fixing the cause so the initial run completes.

Written by the indexing model from the issue text.