From software engineering to context engineering
Why the developer craft is shifting from writing code to managing the context AI works in — and four concrete steps a team can take this week.
Notes
Short pieces on bringing AI into real developer work.
The 5 most common mistakes when bringing AI into a developer workflow
The mistakes teams make most when bringing AI into development: shadow IT, speed without tests, vanity metrics, late governance and a tool instead of a process.
Custom MCP servers — teaching AI to understand your own software
What an MCP server is, why a custom one beats generic prompting, and how to set boundaries so a coding agent gets deterministic, safe access to your system.
Tell me what you need: an app built, a team trained, or agents put into production.