AI Coding Agents and How to Code Them
summary
Live coding builds AI coding agents that run a REACT-style thought→action loop, use Model Context Protocol servers (bash, filesystem, fetch, Brave search) and integrate OpenAI/Anthropic models via langchain and lgraph. A multi-agent supervisor delegates to front-end, back-end and DevOps agents which create files, Docker compose, Git commits and a working conference site, update content from the real website and add Grafana/Prometheus monitoring; site and CI/CD mostly work, Prometheus scraping needs debugging, and agents can post status to Slack and open pull requests.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 1:15Developer time use
- 4:51Design principles
- 12:43Basic agent prototype
- 27:53Model Context Protocol
- 31:22Custom MCP servers
- 42:46Multi-agent LangChain
- 54:49Demo in Encoder IDE
- 1:06:16Monitoring with AI agents
- 1:17:17Future of AI coding
- 1:23:42Q&A
related talks
40mAI Coding Agents and how to code them
48mAI Coding Agents and how to code them
48mAI Coding Agents and How to Code Them
01h 28mAI Coding Agents and how to code them
01h 03mAI Coding Agents and how to code them
26mAI Agents as Your Co-Developer
41mGoing beyond copilot with AI Agents
24mMixing AI Coding in IDEs, CLIs, and Cloud