This course covers practical foundations and advanced techniques for building and orchestrating AI agents across major platforms (Codeex, Claude, Gemini). It combines live demos—parallel browser automation, multi-agent orchestration, and video-to-action pipelines—with design patterns for reliability, self-modifying prompts, and token/context management. The presenter explains setup, trade-offs between speed, cost and quality, and reproducible skills for real projects.
– Core agent loop: observe → think → act, iterate to a clear definition of done.
– Platform setup: step-by-step demos and differences between Codeex/Gemini/Claude.
– Advanced patterns: MCP orchestration, stochastic multi-agent consensus, agent chat rooms, and subagent verification loops.
– Practical controls: self-modifying agents via agents.mmd, prompt contracts, reverse prompting, skills, and context/window management to balance cost and accuracy.
Quotes:
The main strength of AI agents is really their ability to parallelize.
The LLM is just a very small part of what most people consider AI agents.
Agents can learn from YouTube videos—the same medium humans learn from.
Statistics
| Upload date: | 2026-03-08 |
|---|---|
| Likes: | 4699 |
| Comments: | 261 |
| Statistics updated: | 2026-03-30 |
Specification: AI Agents Full Course 2026: Master Agentic AI (2 Hours)
|
