99% of LLM-adjacent software is webslop in a state of perpetually broken churn.

OpenCode for example has not yet figured out "maintain a log of all messages and send that log to the SSE endpoint in the same order to get the next response" and has regular prompt cache misses even with context pruning disabled