Reflection

The first time an agent blows it, the instinct is to reach for a bigger model. The second time, to “prompt harder.” Both miss the point. An agent that goes wrong is usually not a dumb model. It’s a model you handed the wrong shape to work in. The patterns you choose — how the model routes, what tools it can touch, whether it checks its own work — are what separate an agent that seems smart from one that keeps failing in the same predictable way. ...

A year ago this site was an empty scaffold and a promise. A year of building later, there’s a body of work: things shipped, things broken, things learned. This is the honest ledger of that year — what actually compounded, what quietly didn’t, and the one insight I’d reread on day one if I could. What actually compounded The streak, not the sprint. The single biggest lesson, and the least glamorous. Consistent small sessions beat heroic weekends every time, because the streak keeps the context warm and the weekends keep relearning it. I’d heard this for years. I didn’t believe it until I felt the difference. ...