Fixing a duplicated changelog upstream of the parser
Where the defect appears and where it's best fixed are different questions. I have a habit of answering the first and acting as though I have answered the second.
30notes from production work
I build software end to end. I'm Scott Mallinson, a senior software engineer at Cognizant in Barcelona, Spain. I write about API design, type systems, guardrails for AI features, and the maintenance that keeps a codebase navigable.
Where the defect appears and where it's best fixed are different questions. I have a habit of answering the first and acting as though I have answered the second.
I capped my assistant's tool catalogue at three tools per integration and six hundred tokens. Every new capability since has had to become a parameter on an existing tool.
I wrote a 1.5 second latency target into a spec and reasoned from it for weeks. Measured, the real figure was 60 to 110 seconds.
An auth token held in memory and a SQLite file on disk both worked locally and disappeared in production, because my laptop provides durability the serverless platform doesn't.
Two background jobs stopped early and recorded themselves as complete, and once recorded, nothing in the system could reopen them.
Chief's daily brief runs on a small local model. Getting it to behave meant moving a bullet-count limit out of the prompt and into code, then measuring what a repeat penalty setting actually cost in accuracy.