Luna utility

SkillProductivity

Lets your agent review code for correctness, security, and reliability issues.

Available today. Use it from your connected AI after setup.

Connect ahel once, and every AI you use reads what you have installed.

Then ask your AI: use the Luna utility skill

About this capability

Use Luna to watch, grade, or run bounded worker tasks. Use when the user invokes /luna or wants the Luna utility.

What this skill tells your AI

The instructions your AI receives, as published by codejunkie99/codex-self-improving-loop in skills/luna/SKILL.md and read by ahel’s review.

Run the task on gpt-5.6-luna with the Codex CLI:

codex exec -m gpt-5.6-luna -s read-only --skip-git-repo-check --ephemeral -C . "<task>"

For grading a loop run, feed the current prompt.md plus log.md to Luna with the watcher instructions from the self-improving loop. Luna never rewrites prompts; it grades and emits GOAL, KEPT, WASTED, FAILED, NEXT.

Signals

GitHub stars
48
Forks
7
Last commit
Sep 2026
Advanced
Catalog kind
skill
Gateway key
luna
Source
github.com/codejunkie99/codex-self-improving-loop