What problem does it solve?
Editing generated binding files by hand or regenerating them incorrectly leaves the repository silently stale or breaks the build across a dozen language targets. This Skill encodes the exact Alef regeneration workflow, the alef.toml layout, and the core-side Rust edits that break a regen, so changes to packages/* and binding crates stay consistent.
Core Features & Use Cases
- Regeneration Workflow: Runs task alef:generate, task alef:verify, and the e2e task chain in the correct order, with atomic commits of Rust source, alef.toml, and generated output.
- Freshness Verification: Uses alef verify and .alef/ hash caches instead of under-scoped git diffs, covering crates/xberg-py, xberg-node, xberg-ffi, packages/dart/rust, and e2e output.
- Breakage Prevention: Documents the core-side edits that abort a regen, such as missing #[cfg_attr(alef, alef(skip))] on new extractors, #[non_exhaustive] on binding-facing types, and trait return types lacking Default + Serialize + Deserialize for the FFI bridge.
- Use Case: You added a method to a trait with a generated FFI bridge. The Skill tells you the return type needs Default + Serialize + Deserialize, and to verify with cargo check -p xberg-ffi rather than the workspace check.
Quick Start
Ask the assistant to regenerate the Alef bindings after modifying a Rust extractor and verify the output is fresh.