skf-rename-skill

To safely rename an agent skill, you need a transactional workflow that atomically updates all versioned directories, manifest entries, and platform context files. This ensures references remain synchronized and prevents data loss by verifying the new name before removing the old directories.

Updating SKILL.md frontmatter and export manifests during a skill name change requires comprehensive reference updates across metadata.json, context-snippet.md, and provenance-map.json. A copy-verify-delete pattern guarantees all structural references and managed context sections are fully materialized before the original skill is removed.

Yes, you can rename a BMAD agent skill across multiple versioned releases without manual file edits by using an automated rename workflow. It handles directory renames, reference fixes, and platform context rebuilds across every version automatically, requiring zero manual intervention.

Renaming a skill to kebab-case format can automatically update IDE-specific context files like CLAUDE.md and .cursorrules. The process rebuilds platform context files and validates the new name against the agentskills.io specification to ensure directory names match parent skill names.

If a skill rename fails halfway through, transactional safety ensures the original skill remains completely intact. By using a copy-verify-delete pattern, the workflow only removes the old skill directories after the new name is fully materialized and verified, leaving the original untouched upon any failure.

Handling orphaned draft skills with no manifest entries during a rename requires validation guardrails that detect missing export manifests. The workflow checks for name collisions and warns about official source authority mismatches to safely process orphaned drafts without breaking existing references.