AI Agent Engineering Interview, issue 2, Sep 21, 2026
The Skill Overload Trap
Senior AI Engineer interview at Anthropic, and the interviewer asks:
“You gave your coding agent 40 carefully written skills to make it smarter, and task success went DOWN. Each skill description is only a few lines, so it’s not a context-budget problem. What’s actually breaking, and how would you prove it?”
Don’t say: “Too many tokens in the system prompt. Trim the descriptions.”
Why tools fail on selection, not storage, and how to prove instruction hijacking when extra capabilities start destroying agent performance.
The full answer, with the mechanism and the arithmetic, is for paid subscribers on Substack.