fix(projects): handle large synchronization histories

[zoeyrose]

Summary

Project Synchronization Health crashes with jq: Argument list too long when GitHub returns a large synchronization history. Stream the API responses through private temporary files so large histories can be classified normally.

Implementation / behavior

• Merge the two workflow responses from files instead of passing full JSON in command-line arguments.
• Preserve API exit codes, JSON truthiness checks, single-document workflow validation, incident ownership, and recovery behavior.
• Add regressions with responses exceeding 2 MiB and malformed, empty, multiple-document, and failed API responses.

Validation

• Full repository test suite, bin/validate, Bash syntax checks, ShellCheck 0.11.0, and git diff --check passed at 86af35cb2ff4bc388cfe5cc3af79b02cec083ef3.
• Independent review and an independent focused health-check suite passed at the same revision.
• Remote Policy validation, Conventional PR title, and CodeQL checks passed on the reviewed head and current main base.

Limitations / follow-up

The monitored synchronization workflow separately reports pending Pages activation has unexpected live source for atrinik/classic. This fix preserves that unhealthy result; it does not change Pages settings, apply governance, or dispatch production workflows.

[github-actions]

🎉 This PR is included in version 2.9.0 🎉

The release is available on GitHub release

Your semantic-release bot 📦🚀