Super skill
Git-backed version control, rollback, and provenance for your Claude Code & Codex Agent Skills — a safety-gated skill package manager (CLI + plugin).
npx -y skills add sdsrss/super-skill --skill super-skillAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- 28 days oldThe repository was created 28 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Use when the user wants to manage, version, audit, or roll back their Agent Skills, or to review skill candidates mined from past sessions. super-skill is a git-backed package manager for the skills in ~/.agents/skills (and ~/.claude/skills) — it seeds existing skills under version control, explains where each came from, rolls any skill back one command, checks registry integrity, and surfaces recurring task families worth turning into a skill. Triggers include "version my skills", "roll back that skill", "why do I have this skill", "check my skill registry", "what could become a skill".
SKILL.md
2.8 KB, 584 tokens by cl100k_base, as published. Nobody here has run it
super-skill (Codex)
super-skill is a personal package manager for Agent Skills. It does not run
your skills — it manages them: version history, provenance, rollback, integrity
checks, and opportunity mining. It is driven by the host-agnostic super-skill
CLI (install with pipx install super-skill-cli or uv tool install super-skill-cli; the command is super-skill).
Scope is deliberately the package-manager form. The self-learning loop (auto-optimizing skills, external distillation) is a deferred research track and is NOT part of this tool — do not imply it exists.
When to use
- Version / audit existing skills →
super-skill seedimports your skills dir under version control (read-only on the host) →super-skill list/status. - "Why do I have this skill / where did it come from" →
super-skill explain <id>. - A skill regressed, undo it →
super-skill rollback <id> [--to vN]. - Registry health →
super-skill doctor(read-only);super-skill doctor --fixrestores git-recoverable versions and re-materializes drift, then re-verifies. - "What do I keep re-solving that could be a skill" →
super-skill mine→super-skill candidate draft→ review →super-skill candidate approve <id>.
Codex specifics
- Skills live in
~/.agents/skills/<name>/SKILL.md(user-level) or.agents/skills(repo-level). The super-skill registry is the source of truth;~/.agents/skillsis a materialization target it writes atomically — do not hand-edit skills there (doctorflags such edits as host drift androllback/materializeoverwrite them). - Select Codex as the target with
--host codex(dir override:SUPER_SKILL_CODEX_SKILLS, default~/.agents/skills):super-skill seed --host codex,super-skill materialize --host codex,super-skill rollback <id>(re-syncs every host the skill was distributed to). Do NOT repointSUPER_SKILL_HOST_SKILLSat~/.agents/skills— that hijacks the Claude host slot and later writes land in the wrong directory. - Run the CLI via shell and interpret the output for the user.
seed/doctor(no--fix) are read-only; every write path is explicit and reversible.
What ships with it: 1 file
640 B alongside SKILL.md
agents/
- openai.yaml640 B