Best of 2025
Good Vibes and Motivation
My thoughts on software development, life, and more.
Two months after a coding agent turned a failing test green and left an inconsistent API behind, I gave another agent the brief I should have written. It stopped three times, and each stop was the right call.
An agent that returns a diff and a cheerful "Done!" moves the work to the reviewer. What evidence different tasks should return, which checkpoints keep long-running work visible, and why a passing report isn't a release decision.
Autocomplete, pair programming and delegation need different amounts of structure. A five-point readiness check and a short delegation brief turn a vague request into work an agent can finish and a human can review.
Approval checkpoints for coding agents: which actions should pause for a human, what the approver needs to see, and how to match safeguards to risk without making low-risk work painful.
Takeaways from WorldQuant University's Foundations of Disruptive Systems: how system design shapes the spread of disruption and the difficulty of recovery. Single points of failure, the trade-off between efficiency and resilience, causal chain maps, leading indicators and accountability, with parallels for developers and an interactive cascade model.
Why a Codex rollout is an operating-model decision, not a product-adoption decision: find repeated delivery friction, bound one pilot, keep a human accountable, and expand only when both value and safeguard evidence hold up.
How the chat on owenadirah.com answers questions about my work with Mastra and a Neo4j knowledge graph, and why every sentence it says must cite a source or it says "I don't know".
A small shell that composes an Angular remote and a React remote at runtime with Native Federation, and the ownership rules that keep auth, navigation and feature policy from leaking into every app.
Playlists to keep motivated and creative.