
The podcast explores the evolving landscape of AI coding tools, techniques, and terminology. It highlights the shift from older methods like cursor rules to newer concepts such as agents.md, skills, and sub-agents. Scott and Wes discuss various AI coding tools, including text editors with AI features (VS Code, Copilot, Cursor), terminal-based interfaces (Claude Code, OpenCode TUI), and full GUI desktop apps (Codex app, OpenCode app), with a preference for GUI-based tools due to better information density and UI. They also delve into the importance of choosing the right AI model (Codex 5.3, Opus 4.6) for specific coding tasks and the use of agents, sub-agents, skills, and slash commands to enhance coding efficiency. The discussion further covers the use of hooks and plugins for code quality and the role of MCPs in connecting AI with external apps.
Sign in to continue reading, translating and more.
Continue