
		<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
			<channel>
				<title>MikeDaDev&#39;s Blog – stories, insights &amp; ideas</title>
				<link>https://mike.tungdadev.com/blog</link>
				<description>A personal space on the cloud where I document my programming journey, sharing lessons, insights and resources for fellow developers.</description>
				<language>en</language>
				<managingEditor>bthanhtung.ite@gmail.com (MikeDaDev)</managingEditor>
				<webMaster>bthanhtung.ite@gmail.com (MikeDaDev)</webMaster>
				<lastBuildDate>Wed, 18 Nov 2026 00:00:00 GMT</lastBuildDate>
				<atom:link href="https://mike.tungdadev.com/tags/ai-agents/feed.xml" rel="self" type="application/rss+xml"/>
				
		<item>
			<guid>https://mike.tungdadev.com/blog/all-code-is-slop</guid>
			<title>All Code Is Slop: how the Grok Bot team ships 11,000 PRs a month</title>
			<link>https://mike.tungdadev.com/blog/all-code-is-slop</link>
			<description>A summary and close reading of Lauren Tan&#39;s (SpaceX) talk at AI Engineer Code 2026: turn every agent correction into a lint rule, spend tokens on verification instead of reading code, and why automating a loop that produces slop only gives you more slop.</description>
			<pubDate>Wed, 18 Nov 2026 00:00:00 GMT</pubDate>
			<author>bthanhtung.ite@gmail.com (MikeDaDev)</author>
			<category>ai-agents</category><category>clean-code</category><category>testing</category><category>architecture</category><category>ai</category>
		</item>
	
		<item>
			<guid>https://mike.tungdadev.com/blog/ai-replace-developers</guid>
			<title>Will AI replace developers?</title>
			<link>https://mike.tungdadev.com/blog/ai-replace-developers</link>
			<description>Colleagues, family and AI company CEOs all say developers are about to be replaced. This post sets each claim next to the primary data from Stanford, METR, Anthropic, DORA, BLS and Indeed, plus codeaholicguy&#39;s essay, to see what holds up and what is exaggerated.</description>
			<pubDate>Mon, 16 Nov 2026 00:00:00 GMT</pubDate>
			<author>bthanhtung.ite@gmail.com (MikeDaDev)</author>
			<category>career</category><category>ai</category><category>ai-agents</category><category>productivity</category>
		</item>
	
		<item>
			<guid>https://mike.tungdadev.com/blog/agentic-development-life-cycle-adlc</guid>
			<title>ADLC: why your SDLC quietly breaks the day you ship an AI agent</title>
			<link>https://mike.tungdadev.com/blog/agentic-development-life-cycle-adlc</link>
			<description>The Agentic Development Life Cycle from a tech lead&#39;s chair: what changes when behaviour lives in prompts, models and tools rather than code, how to make evals statistically honest, and the eight-stage loop that keeps an agent reliable after launch.</description>
			<pubDate>Sun, 15 Nov 2026 00:00:00 GMT</pubDate>
			<author>bthanhtung.ite@gmail.com (MikeDaDev)</author>
			<category>ai-agents</category><category>testing</category><category>observability</category><category>ai</category>
		</item>
	
		<item>
			<guid>https://mike.tungdadev.com/blog/agentic-ai-architecture</guid>
			<title>Agentic AI architecture: the boring parts that decide whether your agent survives production</title>
			<link>https://mike.tungdadev.com/blog/agentic-ai-architecture</link>
			<description>A tech lead&#39;s map of agentic AI architecture: workflows versus agents, the runtime around the model, tool design and MCP, context and cache economics, durable execution, multi-agent trade-offs, and security patterns that hold up against prompt injection.</description>
			<pubDate>Sat, 14 Nov 2026 00:00:00 GMT</pubDate>
			<author>bthanhtung.ite@gmail.com (MikeDaDev)</author>
			<category>ai-agents</category><category>architecture</category><category>ai</category>
		</item>
	
		<item>
			<guid>https://mike.tungdadev.com/blog/claude-code-12-features</guid>
			<title>Claude Code: 12 features that keep your context clean and your agent on track</title>
			<link>https://mike.tungdadev.com/blog/claude-code-12-features</link>
			<description>Using Claude Code only as a chat box is a waste. Learn 12 features, from CLAUDE.md, plan mode, skills and subagents to hooks, that keep your context clean and the agent under control.</description>
			<pubDate>Sun, 18 Oct 2026 00:00:00 GMT</pubDate>
			<author>bthanhtung.ite@gmail.com (MikeDaDev)</author>
			<category>ai-agents</category><category>productivity</category><category>ai</category>
		</item>
	
		<item>
			<guid>https://mike.tungdadev.com/blog/agentic-rag</guid>
			<title>Agentic RAG: when one search is not enough, and what it costs</title>
			<link>https://mike.tungdadev.com/blog/agentic-rag</link>
			<description>Classic RAG searches once, so it often gets multi-step, multi-source questions wrong. See how to build Agentic RAG, from hybrid search and reranking to a tool-calling loop and cost control.</description>
			<pubDate>Wed, 14 Oct 2026 00:00:00 GMT</pubDate>
			<author>bthanhtung.ite@gmail.com (MikeDaDev)</author>
			<category>rag</category><category>ai-agents</category><category>llm</category><category>ai</category>
		</item>
	
		<item>
			<guid>https://mike.tungdadev.com/blog/herdr-ai-agent-multiplexer</guid>
			<title>What is Herdr? Manage a whole team of AI agents in one window</title>
			<link>https://mike.tungdadev.com/blog/herdr-ai-agent-multiplexer</link>
			<description>Run several AI agents at once and you forget which tab is waiting for approval, and a dropped SSH connection kills the session. Herdr is a Rust multiplexer that knows which agent is waiting, keeps your sessions alive and lets agents orchestrate each other.</description>
			<pubDate>Sat, 03 Oct 2026 00:00:00 GMT</pubDate>
			<author>bthanhtung.ite@gmail.com (MikeDaDev)</author>
			<category>ai-agents</category><category>productivity</category><category>ai</category>
		</item>
	
		<item>
			<guid>https://mike.tungdadev.com/blog/build-ai-employees-agent-playbook</guid>
			<title>An AI workforce for your IT team: 4 agents running on Bash, jq and cron</title>
			<link>https://mike.tungdadev.com/blog/build-ai-employees-agent-playbook</link>
			<description>Reading PRs, collecting Jira tickets and writing reports every morning? Build 4 agents (Dev, PM, QC, DevOps) with Bash and cron, so the reports are already in a folder before you open your laptop.</description>
			<pubDate>Wed, 30 Sep 2026 00:00:00 GMT</pubDate>
			<author>bthanhtung.ite@gmail.com (MikeDaDev)</author>
			<category>ai-agents</category><category>productivity</category><category>devops</category><category>ai</category>
		</item>
	
		<item>
			<guid>https://mike.tungdadev.com/blog/spring-ai-dynamic-tool-calling</guid>
			<title>Dynamic tool calling in Spring AI: send only the tools you need</title>
			<link>https://mike.tungdadev.com/blog/spring-ai-dynamic-tool-calling</link>
			<description>An agent with 100 tools ships a hundred JSON schemas with every request, wasting tokens and inviting wrong calls. Select tools dynamically in Spring AI with a CallAdvisor, a Vector Store and permission filtering.</description>
			<pubDate>Sat, 26 Sep 2026 00:00:00 GMT</pubDate>
			<author>bthanhtung.ite@gmail.com (MikeDaDev)</author>
			<category>ai-agents</category><category>llm</category><category>spring-boot</category>
		</item>
	
		<item>
			<guid>https://mike.tungdadev.com/blog/langgraph-multi-agent</guid>
			<title>LangGraph and stateful multi-agent systems: loops that let agents fix their own mistakes</title>
			<link>https://mike.tungdadev.com/blog/langgraph-multi-agent</link>
			<description>A one-way chain cannot send an agent back to fix a failing test. Use LangGraph to build a state machine with loops, human-in-the-loop, time travel and multi-agent coordination patterns.</description>
			<pubDate>Wed, 02 Sep 2026 00:00:00 GMT</pubDate>
			<author>bthanhtung.ite@gmail.com (MikeDaDev)</author>
			<category>ai-agents</category><category>llm</category><category>ai</category>
		</item>
	
		<item>
			<guid>https://mike.tungdadev.com/blog/n8n-ai-agent</guid>
			<title>AI agent workflows with n8n: RAG, tool calling and running in production</title>
			<link>https://mike.tungdadev.com/blog/n8n-ai-agent</link>
			<description>Building AI workflows where most of the time goes into API glue code? A look at the AI Agent architecture in n8n: LLM, Memory, Vector Store, Tool Calling and what you need before going to production.</description>
			<pubDate>Sat, 25 Jul 2026 00:00:00 GMT</pubDate>
			<author>bthanhtung.ite@gmail.com (MikeDaDev)</author>
			<category>ai-agents</category><category>workflow</category><category>rag</category>
		</item>
	
		<item>
			<guid>https://mike.tungdadev.com/blog/mcp</guid>
			<title>Model Context Protocol (MCP): connecting AI to Jira, databases and internal tools</title>
			<link>https://mike.tungdadev.com/blog/mcp</link>
			<description>Connect AI agents to Jira, databases and internal tools through the Model Context Protocol (MCP): the client-server architecture, how to write an MCP server, and how to secure it.</description>
			<pubDate>Sat, 07 Feb 2026 00:00:00 GMT</pubDate>
			<author>bthanhtung.ite@gmail.com (MikeDaDev)</author>
			<category>ai-agents</category><category>ai</category><category>java</category>
		</item>
	
		<item>
			<guid>https://mike.tungdadev.com/blog/ai-agentic</guid>
			<title>AI Agent: From the ReAct Loop to Production Multi-Agent Systems</title>
			<link>https://mike.tungdadev.com/blog/ai-agentic</link>
			<description>What sets an AI agent apart from a standard Q&amp;A chatbot? Dissecting the LLM brain, memory tiers, tools, the ReAct loop, and how to design production multi-agent systems that do not spiral into infinite loops.</description>
			<pubDate>Sat, 20 Dec 2025 00:00:00 GMT</pubDate>
			<author>bthanhtung.ite@gmail.com (MikeDaDev)</author>
			<category>ai-agents</category><category>ai</category><category>java</category>
		</item>
	
			</channel>
		</rss>
	