Evals before prompts: a small eval harness for an LLM feature

Every LLM feature I have built started the same way: a prompt that worked on the three examples I had open. On the roof report extraction work, the examples I had open turned out to be the easy ones.