AI Coding Agents and how to code them
summary
Shows how to build coding agents that use real tools (ping, bash, Playwright, a file system MCP tool) and a React-style reason+act loop, converts simple functions into MCP servers and hooks them into an MCP client, and composes a supervisor-style multi-agent workflow with LangGraph. Claims that adding tool use plus planning raises a hard benchmark from ~6.3% to ~35%. Live demos include a GitHub repo but the multi-agent demo fails during the talk due to provider (Entropic) outage.
AI-written summary · about
chapters (11)
- 0:00AI coding agents intro
- 3:58What defines an AI agent
- 8:54Agentic patterns for coding
- 10:44Building a baseline LLM agent
- 14:16ReAct framework implementation
- 23:08Standardizing tools with MCP
- 37:28Multi-agent systems
- 41:25Supervisor schema creation
- 50:02AI agents in real-world dev
- 53:19Future of software with AI agents
- 1:01:30Q&A on security
related talks
01h 28mAI Coding Agents and how to code them
48mAI Coding Agents and How to Code Them
40mAI Coding Agents and how to code them
48mAI Coding Agents and how to code them
01h 28mAI Coding Agents and How to Code Them
44mHow Does a Coding Agent Work?
36mPutting AI Agents to Work: A New Era of Open Source Connectivity with MCP
32mProgramming language evolution in the AI era