学习用OpenAI Codex构建多智能体系统，掌握规划、执行、评审等角色分工与上下文管理，打造端到端开发流水线。适合开发者与团队负责人。
原始标题：Build Multi-Agent Systems with OpenAI Codex
Build Multi-Agent Systems with OpenAI Codex 是一门专为开发者、软件工程师及技术团队负责人打造的生产级 AI 工作流实战进阶课程。
全课共包含 15 个章节、64 节课时，并附带完整的提示词模版与检查清单资源包，旨在帮助学员摆脱单一的单次提示词（One-off prompting），建立一套严谨的 AI 协同工程流程。课程核心聚焦于如何将大型软件任务拆分为规划者（Planner）、执行者（Implementer）、评审者（Reviewer）、质检（QA）和文档（Documentation）等不同角色的智能体工序，讲授编写具备上下文自适应与约束管理能力的 Codex 契约提示词，并引入结构化任务板、测试与发布卡门（QA Gates）等工程化机制。通过本课程，学员将掌握如何有效调试智能体运行失败、上下文失效及噪音输出等生产环境常见问题，最终在无需深厚 AI 理论背景的前提下，独立构建出一套涵盖从需求拆解、自动化验证到高标准交付的端到端多智能体开发流水线与仪表盘系统。
Published 7/2026
Created by The AI Orchestrator
MP4 | Video: h264, 1920×1080 | Audio: AAC, 44.1 KHz, 2 Ch
Level: All Levels | Genre: eLearning | Language: English | Duration: 64 Lectures ( 6h 4m ) | Size: 2.7 GB
Build practical Codex workflows with subagents, context management, QA gates, and production-ready handoffs.
What you’ll learn
⚡ Design multi-agent Codex workflows with clear roles, responsibilities, and handoff contracts.
⚡ Break large software tasks into planner, implementer, reviewer, QA, and documentation agent workstreams.
⚡ Write effective Codex prompts that preserve context, constraints, and acceptance criteria.
⚡ Use structured task boards, manifests, and progress trackers to coordinate long-running builds.
⚡ Add testing, review, and release gates so agent-generated changes are verified before shipping.
⚡ Debug failed agent runs, stale context, broken assumptions, and noisy outputs without losing control.
⚡ Package reusable project artifacts, examples, and references for students or teammates.
⚡ Build an end-to-end multi-agent dashboard and handoff workflow using production-style practices.
Description
This course contains the use of artificial intelligence.
Build Multi-Agent Systems with OpenAI Codex is a practical course for developers, technical builders, and AI workflow operators who want to move beyond one-off prompting and design reliable multi-agent software workflows.
The course walks through the full lifecycle of agentic engineering with Codex: defining the problem, splitting work into roles, designing prompt contracts, coordinating subagents, managing context, reviewing outputs, debugging failed runs, and packaging a clean handoff. Instead of treating an AI coding assistant as a magic text box, you will learn how to run it as part of a structured production process.
This is not a course about passive AI theory. It is focused on practical execution: how to tell agents what to do, how to keep them aligned, how to verify their work, and how to turn messy autonomous output into dependable software artifacts.
By the end, you will have a complete operating model for multi-agent Codex projects: from the first project brief through task decomposition, agent orchestration, validation, documentation, and handoff. You can apply these patterns to internal tools, product prototypes, courseware, dashboards, automation projects, or any software workflow where multiple AI-assisted workstreams need to stay coordinated.
If you already use AI coding assistants and want a more disciplined way to plan, execute, review, and ship with them, this course gives you a clear framework and a library of practical examples to adapt to your own projects.
Who this course is for
⭐ Developers and technical builders who want to move beyond single-agent prompting.
⭐ Software engineers adopting AI coding assistants for serious projects and team workflows.
⭐ Technical leads designing agent workflows with planning, review, QA, and handoffs.
⭐ Automation-minded creators building tools, dashboards, courseware, or prototypes with AI agents.
此处内容需要权限查看
会员免费查看
