Codex Astra Orchestrator + Luna Subagents Skill Update
Melvin Vivas · X post · 2026-09-07 · Open on X
Topics: AI Agents, Tool Use & MCP, AI Dev Tools & Productivity · Level: intermediate
Summary
The creator updated his Codex skill, in which a strong model (Astra or Sol) orchestrates and cheaper Luna subagents do the work. The update enforces Luna more strictly for tasks like research and codebase exploration.
Key points
- Pattern: a strong model orchestrates and cheaper models run the subagents.
- Send research and exploration tasks to the cheaper Luna subagents.
- The update makes Codex use Luna more reliably for those tasks.
- Astra or Sol can be the orchestrator.
Resources mentioned
- donvito/codex-astra-luna-orchestrator · repo · github.com · free
The creator's Codex skill for an orchestrator-plus-subagents setup (Astra orchestrator, Luna subagents), configured through config.toml.
Also in: Codex Orchestrator Pattern: Astra/Sol Orchestrator with Luna Subagents (Melvin Vivas on X · notes), Open-source Codex skill: Astra/Sol orchestrator with Luna subagents (Melvin Vivas on X · notes), Cost-Efficient Codex Subagents: Astra/Sol Orchestrator + Luna Workers (Melvin Vivas on X · notes), Codex Orchestrator v0.2.1: GPT-6 Sol orchestrator with Luna subagents (Melvin Vivas on X · notes) and 27 more
Try this
- Pull the updated skill if you use the Astra-Luna setup.
More in AI Agents, Tool Use & MCP
- Cutting Codex Subagent Costs: Luna Orchestrator, Astra Reviewer
- Cutting Token Costs in Codex Orchestrator/Subagent Setups via config.toml
- Inspecting Coding-Agent Traces Live with JSONL Viewer (Codex, Claude Code)
- Learn Subagents in Cursor and Codex
- Codex Subagents: Official Documentation
- Testing Model Combinations for Codex Orchestration