fix: align aube install scripts and cold-cache cleanup

[jdx]

The package-manager benchmarks aim to exclude lifecycle scripts, but aube was invoked without --ignore-scripts. aube 2.4.0 runs builds for its bundled trusted dependencies, so Babylon's install was compiling native modules and running Puppeteer downloads while Bun and pnpm had scripts disabled.

Pass --ignore-scripts to aube's regular install, CI command, and CI lockfile preparation. This also covers the bare command used for process counting:

aube install --ignore-scripts --silent
aube ci --ignore-scripts --silent

Also clear the configured content store returned by aube store path and honor XDG_CACHE_HOME. The old cleanup only removed ~/.cache/aube and ~/.aube-store, leaving the current XDG data store populated across supposedly cold runs. Keep the legacy-directory cleanup for older releases. This correction can increase cold-run times; both changes are needed for comparable workloads.

Validation on Linux x86-64 with aube 2.4.0 and the Babylon fixture at aff8423:

Command	Three wall-time samples	Median
aube install --silent	33.92s, 35.02s, 37.82s	35.02s
aube install --ignore-scripts --silent	28.01s, 28.51s, 28.40s	28.40s

Each untraced run started without node_modules or a lockfile and with a fresh aube metadata cache and content store. The six runs interleaved the two commands and used the public registry; previously downloaded Puppeteer browsers remained cached. All six succeeded with the same 1,694-package graph. Separate strace runs showed 448 execution attempts with the default policy versus 2 with scripts disabled. These measurements validate the mismatch; they do not predict the site's ARM64 results or the combined effect of correcting its store cleanup.

All eight Node tests pass, including new coverage for a configured store path containing spaces, a failed store-path lookup, and preserving the store during metadata-only cleanup. Bash syntax checks and ShellCheck's error-level checks pass for the changed scripts.

AI-assisted — Tool: Codex; model: OpenAI/unavailable; version: unavailable.

[coderabbitai]

Navigate logical layers of code changes, visualize relationships, and explore their blast radius.

Note

Currently processing new changes in this PR. This may take a few minutes, please wait...

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Advanced

Run ID: 68296dcf-53bc-4ddc-ba47-6ba7b3615579

📥 Commits

Reviewing files that changed from the base of the PR and between aff8423 and d7eea33.

📒 Files selected for processing (4)

• scripts/clean-helpers.sh
• scripts/clean-helpers.test.js
• scripts/variations/ci.sh
• scripts/variations/common.sh

 _______________________________________________________________________________________________________
< Start when you're ready. You've been building experience all your life. Don't ignore niggling doubts. >
 -------------------------------------------------------------------------------------------------------
  \
   \   \
        \ /\
        ( )
      .( o ).

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create a new PR

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}