» Tag
production
2 posts6–Jul 12, 2026
Evaluations Are Measuring the Wrong Thing
Most evaluations do not measure the reliability of AI models in production. Here are suggestions for accurate assessment.
What Breaks MCP Servers in Production
A look at real production failures in MCP servers—model-unreadable errors, token rotation logouts, schema key rejections, size limits—and the standard built to fix them.