The ~60-line body is lean and information-dense, assuming Claude knows Cargo/PyPI/Maven; the only explanatory aside (a bare xberg>=1.0.0 excludes rcN per PEP 440) is a non-obvious gotcha that earns its tokens rather than padding.

Provides copy-paste commands (task version:sync, alef sync-versions, python3 scripts/sync_integration_versions.py, task version:bump:*, task version:set, task version:check) plus exact config locations (alef.toml [workspace.sync], VERSION_TARGETS, XBERG_DEP_MANIFESTS, NPM_MANIFESTS) for adding integrations.

The sync is given as an ordered sequence (1. alef sync-versions, 2. integration script) with an explicit validation checkpoint — task version:check "dry-runs both and fails on drift" — plus a Do/Don't checklist, so validation is present and not capped at 2.

No bundle files exist, so this single self-contained document with clearly organized sections (source of truth, propagation, lockstep, Do/Don't) and no nested references fits the simple-skill allowance for a top score.

Names multiple concrete mechanisms — "Cargo.toml is the single source of truth", "task version:sync propagates it to alef-managed binding manifests AND the integrations", and lockstep publishing — matching the anchor for listing several specific concrete actions rather than a partial set.

It answers both what (sync/release mechanics with source-of-truth and propagation) and when via the explicit "Load before..." trigger clause, which is an equivalent of a "Use when" clause and avoids the cap at 2.

The trigger clause "Load before bumping a version, editing the version-sync task, or touching an integration's version/xberg dependency" uses natural developer phrasing a user would actually say, giving good coverage rather than jargon-only terms.

The xberg version-sync niche with project-specific triggers (version-sync task, integration xberg pins) is clearly distinct and unlikely to fire for unrelated skills.