Our Agent Had the Checklist and Ignored It
An AI agent had a 41-check quality checklist but kept making cosmetic edits instead of addressing failures. The fix was application-layer forcing.
Apr 8, 20265 min read0 reactions0 comments
Back to articles Apr 8, 2026 
Jan 21, 2026
Tag archive
#llmbehavior
The Chatbot That Started Arguing With Customers Because I Told It to Be Confident
I added one word to a prompt. One single word. That word turned a helpful customer service bot into a...
Jan 21, 20266 min read0 reactions1 comments