From Findings to Fixes: How SCOPE and Azure Cosmos DB Improved the Agentic Product Experience
Articles & talks
Writing and talks from the community on evaluating how coding agents experience your product, and what it takes to fix what they get wrong.
Articles
Section titled “Articles”- Your work might not need the smartest model (opens in a new tab)
- Behind SPFx Dev Skills: testing what agents know and fixing what they miss (opens in a new tab)
- How to test agent experience changes without shipping them (opens in a new tab)
- How to test agent skills without hitting real APIs (opens in a new tab)
- Building AX evals that actually work (opens in a new tab)
- The hidden variables in your agent eval (opens in a new tab)
- Don't rewrite your CLI for agents (opens in a new tab)
- Not all model upgrades are upgrades (opens in a new tab)
- What AI benchmarks are not telling you (opens in a new tab)
- When the model has never seen your code (opens in a new tab)
- Models don't have preferences, they have context (opens in a new tab)
- Stop overloading your skills (opens in a new tab)
- How to diagnose conflicts between AI agent extensions (opens in a new tab)
- Why AI coding agents keep using deprecated CLIs and SDKs (opens in a new tab)