Draft releases cleanup and versioning strategy

[abienkowski]

Problem

The release workflow auto-creates a draft release with an auto-bumped patch version on every push to main. This has produced 6 draft releases (v0.2.1–v0.2.6), none published, with auto-generated notes and no curated content.

Solution

1. Decide the versioning strategy for the first release (e.g., adopt semver; publish as v0.2.6 or bump to v1.0.0)
2. Delete or archive the stale draft releases (v0.2.1–v0.2.5)
3. Curate release notes for the chosen version (feature highlights, breaking changes, links to docs)
4. Optionally suppress auto-draft creation for trivial commits (e.g., only create release on PR merge, or add manual trigger) to reduce noise

Which implementation(s) would this affect?

• All

Additional context

• Workflow: .github/workflows/release.yml — bump job reads latest v* tag, creates tag + draft release via gh release create --draft --generate-notes
• Tracked by epic First release prep: Quint spec review, test parity, and release pipeline validation #4

[abienkowski]

Summary of work completed (for the record):

• Deleted all 7 stale draft releases (v0.2.1–v0.2.7) and their orphaned git tags (v0.2.1–v0.2.5, v0.2.7); kept v0.2.6 as the auto-bump anchor.
• The stale v0.2.7 draft predated PR feat: gate releases on verification + test-parity fixes #13/docs: correct stale test counts and CLI/verification claims #14 (missing the TS Docker fix and docs polish) and would have shipped broken artifacts, so it was not published as-is.
• v0.2.7 is now published as the first official release, built from current main with valid artifacts (binaries, SBOMs, Cosign signatures).
• v0.2.8 followed immediately, auto-created and auto-published by the pipeline with zero manual steps, confirming the existing auto-bump-patch-on-push-to-main strategy works end-to-end going forward.