Entire CLI Upgrade Issue

I just tried to upgrade our CLI via the entire-upgrade plugin. It shells out to Homebrew to perform the upgrade. But what's weird is at the end, the CLI asked me again "do you want to upgrade", which caused a no-op. I wonder if what's happening is that the final summary is shelling out to entire --version or something, which triggers the upgrade check. Here’s the transcript:

09:37:45 (git) main▲ $ entire upgrade --nightly --yes
Detected Entire CLI (homebrew install):
entire 0.8.43-nightly.202607081321.e009b7501 at /opt/homebrew/bin/entire
git-remote-entire 0.8.43-nightly.202607081321.e009b7501 at /opt/homebrew/bin/git-remote-entire
Latest nightly build is 0.8.43-nightly.202607110651.7cd666280
==> Auto-updating Homebrew...
Adjust how often this is run with `$HOMEBREW_AUTO_UPDATE_SECS` or disable with
`$HOMEBREW_NO_AUTO_UPDATE=1`. Hide these hints with `$HOMEBREW_NO_ENV_HINTS=1` (see `man brew`).
==> Auto-updated Homebrew!
==> Updated Homebrew from 20901df0fa to 08a972812c.
Updated 3 taps (entireio/tap, homebrew/core and homebrew/cask).
==> New Formulae
==> Downloading [https://formulae.brew.sh/api/formula.jws.json](https://formulae.brew.sh/api/formula.jws.json)
\[... elided ...\]

You have 9 outdated formulae and 6 outdated casks installed.

==> Updating Homebrew...
Already up-to-date.
==> Would upgrade 1 outdated package
entireio/tap/entire@nightly 0.8.43-nightly.202607081321.e009b7501 -> 0.8.43-nightly.202607110651.7cd666280
==> Do you want to proceed with the upgrade? [y/n]
==> Upgrading 1 outdated package:
entireio/tap/entire@nightly 0.8.43-nightly.202607081321.e009b7501 -> 0.8.43-nightly.202607110651.7cd666280
==> Fetching downloads for: entireio/tap/entire@nightly
✔︎ Cask entire@nightly (0.8.43-nightly.202607110651.7cd666280) Downloaded 20.2MB/ 20.2MB
==> Upgrading entire@nightly
0.8.43-nightly.202607081321.e009b7501 -> 0.8.43-nightly.202607110651.7cd666280
==> Unlinking Binary '/opt/homebrew/bin/entire'
==> Unlinking Binary '/opt/homebrew/bin/git-remote-entire'
==> Linking Binary 'entire' to '/opt/homebrew/bin/entire'
==> Linking Binary 'git-remote-entire' to '/opt/homebrew/bin/git-remote-entire'
==> Purging files for version 0.8.43-nightly.202607081321.e009b7501 of Cask entire@nightly
🍺 entire@nightly was successfully upgraded!
==> Upgraded 1 outdated package
entireio/tap/entire@nightly 0.8.43-nightly.202607081321.e009b7501 -> 0.8.43-nightly.202607110651.7cd666280
Entire CLI upgrade complete (homebrew install):
entire 0.8.43-nightly.202607110651.7cd666280 installed to /opt/homebrew/bin/entire
git-remote-entire 0.8.43-nightly.202607110651.7cd666280 installed to /opt/homebrew/bin/git-remote-entire

Updating Entire CLI: brew upgrade --yes entire@nightly
✔︎ JSON API packages.arm64_tahoe.jws.json Downloaded 15.2MB/ 15.2MB
Warning: Not upgrading entire@nightly, the latest version is already installed
Update complete. Re-run entire to use the new version.

Root Cause

Root-caused it — your hypothesis is close but the mechanism is different: nothing shells out to entire --version. It's the parent entire process's own post-command version check, running with its stale compiled-in version.

What Happened, Step by Step:

1. upgrade isn't a built-in, so the old binary dispatched to the entire-upgrade plugin. The plugin ran brew and successfully upgraded the binary on disk.
2. When the plugin exits 0, the parent mirrors Cobra's PersistentPostRun and runs the version check. But that parent process is still the old binary, so the version check sees the pre-upgrade version baked in at compile time.
3. Your 24h check cache had expired, so it fetched the latest nightly, saw 202607081321 < 202607110651 and indicated