Agentically optimizing LLM prompt cache TTLs for fun and profit

5 points | by matsur 5 hours ago