آموزش Claude Code 2026: تسلط بر تیم‌های عاملی، MCP و اتوماسیون هوشمند

عنوان دوره : Udemy – Claude Code 2026 : Agents Teams, Subagents, MCP & Skills

سطح : متوسط

زمان کل دوره : 2 ساعت و 35 دقیقه

تعداد فايل ها : 27 (1440p)

سال آموزش : 2026

فایل تمرینی : دارد: HTML (1)

مربی دوره : Shan Singh

نرم افزار : Claude

زیرنویس انگلیسی : دارد

Udemy – Claude Code 2026 : Agents Teams, Subagents, MCP & Skills

Master Claude Code in 2026 — from your first terminal command to

production-grade AI agents built with Agent Teams, MCP servers, Skills,

Hooks, Scheduling and Plugins.

Claude Code is the most-loved AI coding tool of 2026, and it is no longer

“AI chat in a terminal.” It is a layered agentic system — and most

developers use maybe 20% of it.

This course covers the other 80%.

You will not just learn what Skills, MCP, Subagents and Agent Teams are.

You will build each one, from scratch, on real files and a real SQLite

database, and see exactly when to reach for which.

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

WHAT MAKES THIS COURSE DIFFERENT

Most Claude Code courses stop at CLAUDE md and a couple of MCP servers.

This one goes all the way to the ceiling:

– Agent Teams — the newest primitive. Lead agents, teammates, task lists

and mailboxes. Why subagents cannot talk to each other, and what to do

about it. Includes a live parallel review of a real e-commerce SQLite

database with three teammates working at once.

– Scheduling — turn Claude Code from a tool you open into an autonomous

agent that runs on its own. Almost no course covers this.

– Hooks — the seven lifecycle events , auto-

formatting Python after every edit, and blocking dangerous bash

commands before they run.

– Plugins & Marketplaces — package everything you have built into a

distributable plugin, then publish your own marketplace on GitHub.

WHAT YOU WILL LEARN

Foundations

– How Claude Code works internally, and how agentic coding differs from

traditional coding

– Claude AI vs Claude Code vs Claude Cowork — which tool for which job

– Opus vs Sonnet vs Haiku, and when each is worth the tokens

– Terminal setup and VS Code setup, side by side

Control & Context

– Slash commands, built-ins, sessions and resume

– How the context window actually works, what it is made of, and auto-

compaction

– CLAUDE md — three ways to create one, how to write it, how to test it

Extending Claude

– Skills: YAML front-matter, markdown body, how Claude loads them on

demand

– Build your own Data-Extraction Skill and Data-Migration Skill

– Install Skills from the marketplace

Connecting Claude to the World

– What MCP is and why tool integration was broken before it

– Database MCP — query SQLite and Postgres without writing SQL

– Playwright MCP — browser automation

– GitHub MCP — let Claude manage your repositories

– Tools, permissions and the trust model

– The 7 MCP servers every developer should know

Agents

– Built-in subagents: Explore, Plan, General-Purpose

– Custom subagents — where they live, how to build a Code Reviewer and an

Orchestrator

– Skills vs Subagents vs Hooks — a clear decision framework

– Agent Teams, enabled and demoed end to end

Automation

– Hooks at every stage of the session lifecycle

– Hourly scheduled agents

– Building and publishing your own plugin and marketplace

WHO THIS COURSE IS FOR

– Developers who want to move past autocomplete and start engineering

with AI

– Data analysts, data scientists and data engineers automating SQL,

pandas and reporting workflows

– Technical founders shipping faster with a small team

– Anyone who has used Claude Code casually and knows they are leaving

most of it unused

REQUIREMENTS

– Basic programming knowledge in any language

– A computer running Windows, macOS or Linux

– A Claude account (free tier is enough to follow along)

Every concept is taught with a live demo or a hands-on task. No slides

without a terminal behind them.

Enroll now and build the Claude Code setup that actually works.

What you’ll learn:
– Use Claude Code efficiently to write, debug, and build applications with AI
– Master Agentic Coding & Prompt Engineering
– Learn agentic engineering using Claude Code
– Understand and implement sub-agents, skills, MCPs, hooks, and plugins
– Plan, architect, and ship complete projects using Claude Code

Requirements:
– Basic programming knowledge of Python will be Enough !
– An active Claude Code subscription (or Anthropic API key)

Who this course is for:
– Developers (beginner to experienced)
– Business leaders, product managers, and team leads
– Aspiring Data Analyst & data Scientist