engineering / skill

Codex Sol/Astra Orchestrator + Luna Subagents

Install a Codex profile with a GPT-6 Astra, GPT-6.1 Sol or GPT-6 Luna orchestrator, GPT-6 Luna subagents and a reviewer.

Repository owner donvito

Not rated yet
0 comments

Like 0 Save

Published 11 Oct 2026Listing updated 12 Oct 2026

At a glance

Purpose: Install a Codex profile with a GPT-6 Astra, GPT-6.1 Sol or GPT-6 Luna orchestrator, GPT-6 Luna subagents and a reviewer.
Runs in: Local agent with shell
Operating systems: macOS, Linux, Windows
Needs: Runtime: Not stated; Needs internet: Not stated; Needs accounts or API keys beyond your agent: Not stated
Files and data: Reads files: Yes; Writes files: Yes; Sends data to other services: Not stated
Tool type: Skill
Delivery: From GitHub
Version: Latest on the main branch
Access: Code on GitHub
Compatibility: According to the README: Codex CLI, on macOS and Linux (setup.sh) and on Windows with PowerShell or PowerShell 7 (setup.ps1). The profiles use GPT-6 Astra, GPT-6.1 Sol and GPT-6 Luna models.
External costs: The repository is Apache-2.0 licensed and states no pricing. The README links a guide on token usage and measurement. Your coding agent or model provider may charge for usage.
Permissions: The skills and subagents act through Codex with the access you give them. The README's example asks a worker to implement an endpoint, and its diagram shows the root agent integrating the work into the repository. According to the README, the installer copies configuration to .codex/, a skill to .agents/ and instructions to AGENTS.md in the target repository you name. It updates existing component files only after confirmation and keeps existing AGENTS.md content. The README's Windows PowerShell command uses -ExecutionPolicy Bypass. For personal setup you copy files into ~/.codex/agents/ and ~/.agents/skills/ and merge into ~/.codex/config.toml.
Usage terms: Apache License 2.0, as stated in the repository. See its LICENSE file for the exact terms.

Code on GitHub

donvito/codex-astra-luna-orchestrator

Apache-2.0

Python1,705 stars135 forksLast pushed 1 Oct 2026 UTC

Use Astra or Sol as orchestrator and Luna for subagents in Codex

Public GitHub data fetched 13 Oct 2026 UTC · 0 days old

Skill folders in this repository (6)

Folders that contain a SKILL.md file, from GitHub's file list on 13 Oct 2026 UTC.

View on GitHub ↗

What it does

This repository provides Codex profiles for an orchestrator, execution subagents and a reviewer. According to the README, the root can use GPT-6 Astra, GPT-6.1 Sol or GPT-6 Luna; explorer, worker, tester and researcher roles use Luna, and reviewers use Astra in Pro/Plus profiles or Sol in Sol profiles.

The README lists pro, plus, pro-max-2-subagents, plus-max-2-subagents, GPT6-SolMax-LunaMax and GPT6-SolMedium-LunaMax. The README's profile table specifies root models, reasoning levels and concurrent subagents. The README says each profile contains Codex configuration, agent TOML files and an astra-orchestrator skill.

The README describes macOS/Linux and Windows PowerShell installers that copy the selected components into an existing target repository. The README says to launch Codex there and invoke $astra-orchestrator explicitly, or let Codex select it for complex work. The README links guides for manual configuration, iteration, coding workflows and token usage.

Who it is for

Codex users who want a preset orchestrator, subagent and reviewer setup for a project or for their personal Codex configuration.

Example tasks

Run the installer for your platform and choose one of six profiles for a target repository.
Start Codex in the target repository and invoke $astra-orchestrator for complex work.
Ask the skill to use the explorer to map a path, a worker to implement it, and the tester and reviewer on the result.
Choose a profile that runs 2 or 4 concurrent subagents.
Set up the roles for personal use by copying files into ~/.codex/agents/ and ~/.agents/skills/ and merging config.toml into ~/.codex/config.toml.

Getting started

The README says to clone the repository, enter its folder and run the installer for your platform. On macOS and Linux that is ./setup.sh. On Windows PowerShell it is powershell -ExecutionPolicy Bypass -File .\setup.ps1, and on PowerShell 7 it is pwsh -File .\setup.ps1. When prompted, enter an existing target repository other than this one, choose a profile by number or name (Enter selects Pro), and confirm which components to install. Then run codex in the target repository. The README also describes manual and global setup. The README has the exact commands.

Limitations

Needs Codex CLI. The README says Codex loads project-scoped configuration only for trusted projects. The installer asks for an existing target repository other than this one. Setup scripts exist for macOS and Linux (setup.sh) and for Windows PowerShell or PowerShell 7 (setup.ps1). The README says the profiles use GPT-6 Astra, GPT-6.1 Sol and GPT-6 Luna, with named roles pinning their models and reasoning levels. For a personal setup you merge config.toml into ~/.codex/config.toml yourself and must not overwrite other Codex settings. The repository can change at any time; check its README before you install.

Data practices

The README does not describe data handling. What is sent where depends on your coding agent and model provider.