资讯：Claude Fable 5.1 / Mythos 5.1 News: Claude Fable 5.1 / Mythos 5.1

这条用法在做什么What this use does

官方 2026-09：Anthropic 发布 Claude Fable 5.1（通用）与 Mythos 5.1（受信访问）。同一底层模型、不同防护等级。定位编码、知识工作与长任务/Agent，并展示科研侧早期例子（蛋白结合、金星高程图、基因组模型加速）。官方称下调 cache read 价。本条是相邻 AI 资讯，不是 Grok Bot 用法。 Official, Sept 2026: Anthropic launched Claude Fable 5.1 (generally available) and Mythos 5.1 (trusted access)—same underlying model, different safeguard levels. Pitch is coding, knowledge work, and long-running / agentic tasks, plus early research demos (protein binders, Venus DEM, genomics speedups). Anthropic says cache-read prices are cut. Adjacent AI news — not a Grok Bot playbook.

收录理由Why we listed it 主源是 Anthropic 公告；X 视频是官方同帖的发布预告片（Alex Albert / Research Product Management，抽象动画），不是实验室结果界面。基准分、蛋白命中率、金星图精度、省钱比例均为官方自述，本库未独立核实。收录为广场可对照的 AI 资讯，不是可开跑的 Grok Bot 路径。 Primary source is Anthropic’s post; the X video is the official launch trailer on the same post (Alex Albert / Research Product Management, plus abstract animation) — not a lab-result UI. Benchmarks, binder hit rates, Venus map accuracy, and savings figures are Anthropic’s claims — unverified here. Catalogued as AI news for the plaza, not a runnable Grok Bot path.

上手路径How to start

1. 先跑一次安全范围Run a safe first pass

   先在可撤回、不外发的范围内试跑一遍，确认输出可用，再考虑扩大权限。 Try it first in a reversible, non-outbound scope. Confirm the output, then widen permissions.

2. 存 SkillSave a Skill

   把跑通的步骤存成 Skill，下次同类任务直接复用，而不是重新口述。 Save the working steps as a Skill so the next similar job reuses them instead of being re-explained.

3. 再上 RoutineThen add a Routine

   确认输出可靠后，再考虑定时或事件触发；例行默认仍应停在待审。 After output is reliable, consider a schedule or event trigger. Routines should still stop at review by default.