Gisting: Compressing LLM Agent context to ↑ throughput and ↓ costshopify.engineering3 points·mellosouls··0 commentsOpen articleSaveView on HN