Developer blog
Notes for building with intent.
Practical writing about agents, context, interface decisions, and the human moments that make AI workflows useful.
Agents
Designing a useful review loop
A small set of checkpoints can make autonomous work easier to trust and easier to improve.
Context
The brief should travel with the work
Keep the original intent visible while ideas become tasks, drafts, and decisions.
Workspace
From prompt to a shareable outcome
A practical model for moving through planning, execution, and review without losing the thread.
Patterns
Make automation legible
Clear status, explicit handoffs, and human control make connected tools feel dependable.
Have a perspective?
Share the workflow you’re learning from.
Good product thinking gets stronger when it is made visible and discussed with other builders.
Visit the community