Kimi-K3 solved 94% of workflows once. Only 13% survived every run
New evidence separates occasional agent capability from dependable operation, while applied teams show where task choice, model training, and runtime retrieval each belong.
Topic · 2 articles
Reliability Engineering
The latest analysis on Reliability Engineering, across Daily Pulses and Weekly Reviews.
October 2026
September 2026
OpenAI's four-model shutdown hides three migrations behind one replacement
Four legacy model IDs point to GPT-5.6 Terra, but their callers do not share an interface, behavior contract, or cost baseline. Treating the change as a string replacement can hide the real cutover risk.