No red flags found in any of the 11 categories — no credential harvesting, no data exfiltration, no curl-pipe-shell installer. Scanned against the SlowMist agent-security taxonomy, refreshed every 8 hours. Full audit →
by irons163 · Codex Skill · ★ 52
Last updated: · Indexed by AgentSkillsHub · Auto-synced every 8h
🔒 Is three-tier-agent-orchestrator safe to install? View the security audit →
Three-Tier Agent Orchestrator 本 README 以中文為主，英文版請見文末。 一個 Codex skill，讓目前主線程承擔 GPT-6 Astra 的總指揮角色：理解目標、拆分任務、做架構決策、檢查結果並整合輸出；並依工作性質把獨立子任務交給 GPT-5.6 Sol Medium 與 GPT-5.6 Luna Max。 角色分工 主要行為 目前主線程直接承擔 Astra 的統籌角色；skill 不自行判定、切換或要求切換主模型。 委派前確認 GPT-5.6 Sol Medium 與 GPT-5.6 Luna Max subagent 是否可用。 啟動 subagent 時隔離主線程歷史：V2 使用 ，V1 使用 。 V2 的 subagent 名稱依實際參數使用固定結尾： 或 。 任一必要模型或指定 reasoning 不可用時，完整停止工作，不以其他模型或較低 reasoning 替代。 能力齊全後，使用明確的目標、範圍、完成條件與驗證方式委派子任務。 由 GPT-6 Astra 檢查重要 diff、測試與證據，並負責最終整合。 必要條件 Codex 支援 skills 與 subagents，且 subagent 工具可直接指定模型與 reasoning effort。 建議主線程選用 ，但這不是 skill 的 hard-stop 條件。 Sol Medium subagent 使用 與 reasoning effort。 Luna Max subagent 使用 與 reasoning effort。 不需要建立額外的 custom agent 設定檔。Skill 每次啟動 subagent 時都會直接傳入指定的模型、reasoning effort 與目前 schema 支援的上下文隔離欄位；實際 subagent 啟動結果才是能力判準。 V2 若支援 ，名稱會使用可辨識的工作前綴與固定角色結尾，例如 或 。Skill 會在啟動前確認名稱結尾與實際 model／reasoning 一致；V1 不傳入此欄位。 安裝 在 macOS 或 Linux 執行： 若目前 task 沒有重
| Stars | 52 |
| Forks | 4 |
| Category | Codex Skill |
| Quality Score | 67.119610209852/100 |
| Last Updated | 2026-09-23 |
| Created | 2026-08-03 |
| Platforms | codex |
| Est. Tokens | ~3k |
These tools work well together with three-tier-agent-orchestrator for enhanced workflows:
Looking for a three-tier-agent-orchestrator alternative? If you're comparing three-tier-agent-orchestrator with other codex skill tools, these 6 projects are the closest alternatives on Agent Skills Hub — ranked by topic overlap, star count, and community traction.
Define task-specific AI sub-agents in Markdown for any MCP-compatible tool.
Multi-agent orchestration extension for Gemini CLI — 22 specialists, parallel subagents, persistent sessions,
Cross-LLM sub-agent orchestration as an Agent Skills. Route tasks to Codex, Claude Code, Grok, GLM, Kimi, Curs
Cognitive architecture for AI-augmented software development. Specialized agents, structured workflows, and mu
The governed runtime for Python agents you can let act: policy on every action, durable runs, sandboxes, budge
Local-first coordination for human and agent work: durable work, decisions, dispatches, evidence, and prompt-f
Explore other popular codex skill tools:
three-tier-agent-orchestrator is Codex 三層多代理協作 Skill：由 GPT-6 Astra 統籌，GPT-6 Sol Max 與 GPT-6 Luna Max 分工執行。. It is categorized as a Codex Skill with 52 GitHub stars.
You can find installation instructions and usage details in the three-tier-agent-orchestrator GitHub repository at github.com/irons163/three-tier-agent-orchestrator. The project has 52 stars and 4 forks, indicating an active community.
The top alternatives to three-tier-agent-orchestrator on Agent Skills Hub include sub-agents-mcp, maestro-gemini, sub-agents-skills. Each offers a different approach to the same problem space — compare them side-by-side by stars, quality score, and community activity.
Grades come from a rule-based scan built on the SlowMist agent-security taxonomy, covering 11 red-flag categories including credential harvesting, data exfiltration, and curl | sh installers. It is a first-layer scan, not a manual audit — we say so rather than overstate it.
The scale of the problem is documented independently: Liu et al. (2026), in a study of 31,132 agent skills, report that 26.1% contain security vulnerabilities. Our own full-catalog census is published as a citable open dataset.
Sources & who's responsible: