Amp Librarian 提速约 3 倍，成本降低 43%
A Faster Librarian - Amp
Amp Librarian 在质量不变的情况下提速约 3 倍、成本降低 43%，现改用无推理的 GPT-5.5、websocket 模式和新版系统提示词。每轮并行工具调用从约 3 次增至约 8 次，搜索由约 15 轮缩至约 5 轮；内部测试耗时从 2 分钟降至 40 秒，成本从 $1.08 降至 $0.47。
正文
当前显示原文，中文全文译文尚未就绪。
The Librarian is now ~3x faster and 43% cheaper, with the same quality.
It now runs on GPT-5.5 (no reasoning) with websocket mode and an updated system prompt that encourages more parallel exploration. The Librarian fires ~8 tool calls in parallel per turn, up from ~3 with Sonnet, and wraps up a search in ~5 turns instead of ~15.
In our internal eval, about a quarter of that speedup comes from OpenAI's websocket mode and the rest from switching to GPT-5.5 with no reasoning:
Here's a comparison:
How does Kubernetes' HorizontalPodAutoscaler handle missing pod metrics when scaling down — does it assume missing pods are at 100% of their resource requests, or 100% of the target utilization? Cite the function and logic in the source.
Sonnet 4.6 (left) took 2 minutes and cost $1.08, while GPT-5.5 (right) took 40 seconds and cost just $0.47.
The Librarian
A subagent for searching code on GitHub