菜单
已展示 11 / 11 个已收集 Skill。
Personal Taiji v5 workflow overlay for yixiali. Use together with taiji-basic-platform-skills when creating or starting yixiali RL/GRPO or SFT training tasks, applying yixiali naming and storage defaults, discovering checkpoints, automatically creating model…
原文语言：英语
使用 YAML 配置和统一 CLI 操作 yixiali 的可复现太极 Pipeline。用于选择或填写已登记的 SFT A20B Midtrain、SFT A20B OPD-v1、RL HY3.1 A3B 模板，执行离线校验与计划，提交或恢复训练、模型卡、模型转换、推理部署、多个评估、Excel 结果导出和收尾阶段，以及查看前后台 Runner 与逐 checkpoint 状态；也用于用户明确要求时开发、修复或扩展 adapter。
在太极平台部署 / 关闭 / 调用混元模型在线推理服务。当用户说"部署 <模型名>""把 xxx 模型部署起来""关掉 <服务>""缩容到 0""这个服务怎么请求""测一下服务通不通"时使用。覆盖三件事：克隆式部署（含必须补建服务组这一坑）、缩容释放 GPU、OpenAI 兼容的最小调用示例。
Initialize or align repository-level AI collaboration scaffolding with AGENT.md, a delegating CLAUDE.md compatibility stub, and .agent/. Use when a user asks to bootstrap a new repository, onboard an existing repository, migrate Claude conventions, create…
原文语言：英语
Curate a daily (or date-range) personalized paper digest from arXiv + HuggingFace. Fetches new & cross-listed papers, reads their abstracts, matches them against the user's preference topics (Agent / World Model / post-training / ...), and writes a…
原文语言：英语
Turn a local Markdown file into a Feishu (Lark) Docx cloud document and push the link to a user via a private bot DM. Uses a Feishu self-built app (tenant_access_token) with the Drive import API and the im/v1 message API. TRIGGER when: user wants to send/push…
原文语言：英语
Convert PDF files to Markdown using pymupdf4llm. TRIGGER when: user asks to convert a PDF to markdown, extract text from PDF, or read/parse a PDF into markdown format. DO NOT TRIGGER when: user just wants to read an existing paper.md or work with…
原文语言：英语
Extract, browse, and download figures from arXiv papers using their HTML rendering. Use this skill whenever the user wants to get figures/images/diagrams from an arXiv paper, extract paper illustrations, download paper visuals, grab charts or plots from a…
原文语言：英语
Read and work with local academic paper folders exported by ZoFiles from Zotero. TRIGGER when: user asks to read/summarize/cite/review/compare a paper that is already downloaded locally, references a local paper folder or path, works with…
原文语言：英语
Import arXiv papers into Zotero with duplicate detection and batch support. TRIGGER when: user asks to import papers, add arxiv papers to zotero, batch import papers, check for duplicate papers in zotero, mentions importing by arXiv ID, or wants to list…
原文语言：英语
Search, browse, and read academic papers online using HuggingFace Papers API and arxiv2md. TRIGGER when: user asks to search for papers online, find recent papers, look up papers on a topic, search arXiv/HuggingFace for papers, check what's new in a research…
原文语言：英语