Google ADK + A2A + MCP CRASH COURSE - The missing guide for building with its LATEST AI agent stack
🧠 Get the ADK + A2A Builder Pack ► https://gum.co/u/r4i7unej
Includes source code, test suite, cheat sheets, .cursor/rules, and more — built for this episode.
🤫 Join the AI Oriented Insiders Club — it's FREE ► https://insiders.aioriented.dev
☕ Support the channel
► https://www.buymeacoffee.com/aioriented
In this crash course, I walk you through Google’s latest AI agent stack — combining the Google Agent Development Kit (ADK), MCP, and the Agent2Agent (A2A) protocol to build a modular multi-agent system with real-world integrations.
✅ Build A2A-compatible agents for Notion + ElevenLabs
✅ Create a coordinating host agent using ADK
✅ Integrate subagents using A2A (Agent-to-Agent protocol)
✅ Build a custom Streamlit UI to run it all
✅ Add .cursor/rules, test coverage, and reusable tools
✅ Download the Builder Pack to follow along
🔗 Full Source Code on GitHub
https://www.github.com/chongdashu/adk-mcp-a2a-crash-course
This is the missing guide for developers building with Google’s open agent framework. Whether you're scaling personal tools or building production agents, this stack gives you full control over orchestration, voice, and interfaces.
Chapters:
00:00 Welcome
03:53 Setup: python 3.13 + virtualenv
06:12 Setup: Package dependencies
07:36 Notion MCP
09:42 Anatomy of an A2A ADK Agent
12:48 Agent 1: Notion
13:20 __init__.py (Notion Agent)
13:38 prompt.py (Notion Agent)
14:52 agent.py (Notion Agent)
18:59 custom_adk_patches.py
21:20 Agent 2: ElevenLabs
22:16 __init__.py (ElevenLabs Agent)
22:19 prompt.py (ElevenLabs Agent)
22:52 agent.py (ElevenLabs Agent)
29:11 From ADK to A2A
32:13 A2A ElevenLabsAgent
33:15 agent_executor.py
35:39 __main__.py
37:08 A2A Agent Card
42:19 agent_executor.py implementation
52:49 test_client.py
56:29 A2A Notion Agent
59:16 Combining ADK + A2A
01:00:25 ADK sub agents vs A2A worker agents
01:03:02 Host Agent
01:05:02 agent.py
01:05:32 Host tools.py
01:06:25 remote_connections.py
01:07:56 ui/app.py (Embedded Host)
01:10:28 Streamlit App
01:13:39 Host __main__.py
01:14:34 Host agent_executor.py
01:15:19 Host test_client.py
01:18:13 A2A Streamlit App
01:20:39 Conclusion
01:21:39 Builder Pack Resources
01:21:45 Resource 1: Cheat Sheet
01:21:51 Resource 2: Lesson Plan
01:22:02 Resource 3: pytest Test Suite
01:22:48 Resource 4: .cursor/rules for ADK & A2A
01:24:22 Outro
🧠 ADK Made Simple Playlist
► https://www.youtube.com/playlist?list=PLWUH7ke1DYK98Di2FF8Ux3IX6qG-mZ3A7
🎮 Vibe Code Games in 12 Minutes
► https://youtube.com/playlist?list=PLWUH7ke1DYK8pBJN331-jaQ_ynq-qBXMs
🕹️ Website
► https://www.aioriented.dev
🐦 Follow me on X
► https://www.x.com/chongdashu
#mcp #ai #aiagents #CursorMCP #Automation #CursorAI #modelcontextprotocol #claude #automation#windsurfai #vibecoding #agentdevelopmentkit #adk #googleadk #agent2agent #a2a
Видео Google ADK + A2A + MCP CRASH COURSE - The missing guide for building with its LATEST AI agent stack канала AI Oriented Dev
Includes source code, test suite, cheat sheets, .cursor/rules, and more — built for this episode.
🤫 Join the AI Oriented Insiders Club — it's FREE ► https://insiders.aioriented.dev
☕ Support the channel
► https://www.buymeacoffee.com/aioriented
In this crash course, I walk you through Google’s latest AI agent stack — combining the Google Agent Development Kit (ADK), MCP, and the Agent2Agent (A2A) protocol to build a modular multi-agent system with real-world integrations.
✅ Build A2A-compatible agents for Notion + ElevenLabs
✅ Create a coordinating host agent using ADK
✅ Integrate subagents using A2A (Agent-to-Agent protocol)
✅ Build a custom Streamlit UI to run it all
✅ Add .cursor/rules, test coverage, and reusable tools
✅ Download the Builder Pack to follow along
🔗 Full Source Code on GitHub
https://www.github.com/chongdashu/adk-mcp-a2a-crash-course
This is the missing guide for developers building with Google’s open agent framework. Whether you're scaling personal tools or building production agents, this stack gives you full control over orchestration, voice, and interfaces.
Chapters:
00:00 Welcome
03:53 Setup: python 3.13 + virtualenv
06:12 Setup: Package dependencies
07:36 Notion MCP
09:42 Anatomy of an A2A ADK Agent
12:48 Agent 1: Notion
13:20 __init__.py (Notion Agent)
13:38 prompt.py (Notion Agent)
14:52 agent.py (Notion Agent)
18:59 custom_adk_patches.py
21:20 Agent 2: ElevenLabs
22:16 __init__.py (ElevenLabs Agent)
22:19 prompt.py (ElevenLabs Agent)
22:52 agent.py (ElevenLabs Agent)
29:11 From ADK to A2A
32:13 A2A ElevenLabsAgent
33:15 agent_executor.py
35:39 __main__.py
37:08 A2A Agent Card
42:19 agent_executor.py implementation
52:49 test_client.py
56:29 A2A Notion Agent
59:16 Combining ADK + A2A
01:00:25 ADK sub agents vs A2A worker agents
01:03:02 Host Agent
01:05:02 agent.py
01:05:32 Host tools.py
01:06:25 remote_connections.py
01:07:56 ui/app.py (Embedded Host)
01:10:28 Streamlit App
01:13:39 Host __main__.py
01:14:34 Host agent_executor.py
01:15:19 Host test_client.py
01:18:13 A2A Streamlit App
01:20:39 Conclusion
01:21:39 Builder Pack Resources
01:21:45 Resource 1: Cheat Sheet
01:21:51 Resource 2: Lesson Plan
01:22:02 Resource 3: pytest Test Suite
01:22:48 Resource 4: .cursor/rules for ADK & A2A
01:24:22 Outro
🧠 ADK Made Simple Playlist
► https://www.youtube.com/playlist?list=PLWUH7ke1DYK98Di2FF8Ux3IX6qG-mZ3A7
🎮 Vibe Code Games in 12 Minutes
► https://youtube.com/playlist?list=PLWUH7ke1DYK8pBJN331-jaQ_ynq-qBXMs
🕹️ Website
► https://www.aioriented.dev
🐦 Follow me on X
► https://www.x.com/chongdashu
#mcp #ai #aiagents #CursorMCP #Automation #CursorAI #modelcontextprotocol #claude #automation#windsurfai #vibecoding #agentdevelopmentkit #adk #googleadk #agent2agent #a2a
Видео Google ADK + A2A + MCP CRASH COURSE - The missing guide for building with its LATEST AI agent stack канала AI Oriented Dev
adk google adk mcp modelconte agent development kit adk tutorial a2a protocol agent-to-agent fastapi tutorial postman tutorial ai agent framework build ai agents autonomous agents ai automation mcp server multi-context protocol ai workflows streamlit ui google io 2025 llm orchestration ai infrastructure open source ai stack ai agent api adk a2a build fastapi agent google ai tools adk made simple
Комментарии отсутствуют
Информация о видео
14 июня 2025 г. 20:01:19
01:24:55
Другие видео канала
Let's DESIGN & BUILD PONG with DEEPSEEK + CLAUDE | 12-Min Game Dev Challenge | Ep 3
WINDSURF + CLAUDE SONNET 3.7 to design and build a 3D GAME VILLAGE SCENE #gamedev #aigamedev
Goodbye, Cursor? Using GitHub Copilot AGENT MODE to Build 3D BREAKOUT in 12 MINUTES | Ep. 4
Unreal Engine AI LLM Chat in Editor - WIP #2 - NPC Following Player
Riding WINDSURF WAVE 3 to BUILD a 3D SOKOBAN clone in 12 MINS | Turbo Mode | Ep. 6
🚠 I made ANOTHER 3D GAME while ON THE GO with my PHONE 🕹️ #REPLIT #CLAUDE
From localhost to production — deploy ANY Google ADK agent with this SIMPLE Cloud Run setup
Unreal Engine AI LLM Chat in Editor - Video Log #1 - NPC Following Player
#REPLIT for a 3D SNOWY CRUISE GAME while ON THE MOVE
This might be the most underrated AI dev stack right now... MCP + ADK is a game changer (Full Guide)
Go beyond ADK Web... Build a Custom Frontend for your ADK agents (Full Tutorial)
Forget MCP... don't sleep on the Google Agent Development Kit (ADK) - Full tutorial
I Was Using Claude Code WRONG Until I Learned These 8 Hacks (Steal them... ESPECIALLY 6 & 7!)
A FAST Guide to Context Engineering with Claude Code: Automate Better Git Commits
Claude Code UNLOCKED: The secret workflow Anthropic doesn't want you to know (Inc. Kimi K2 + Groq)
MCP Server to 10x Your Vibe Coding: Auto Log Capture, Error Fixing & Screenshot in Cursor, Windsurf
Google is ALL IN on AI Agents — Build Yours with ADK, A2A, MCP + FastAPI (Step-by-Step)
MCP Made SIMPLE: Your FIRST Hello World MCP Server. Works for CURSOR & WINDSURF.
The ONLY Supabase MCP Server you'll ever need for vibe coding. Cursor vs Windsurf (Windows & Mac)
Vibe code in Unreal Engine 5 with this MCP | Blueprints+Scripting | Cursor, Windsurf, Claude Desktop
