仓库/Skill 库

目标：将 State Space Models（SSM）/ Mamba 用于医学时间序列分析；总体研究问题：Mamba 模型能否提升基于步态信号检测帕金森病的分类模型性能: To use State Space Models (SSM)/ Mamba for medical time series analysis General research question: Can Mamba models improve the performance of classification models trained to detect Parkinson's disease using gait signals.

本仓库收录了我关于超透镜（metalens）的研究日志与学习笔记，主要用作个人记录空间，用于整理讲义、代码与文献综述；任何对平面光学感兴趣的人也可使用。This repository contains my research logs and study notes on metalenses. It serves primarily as a personal bookkeeping space to document lecture notes, code, and literature reviews, though anyone interested in flat optics is welcome to use these resources.

SLM-PICO-Screener 是用于自动化系统综述筛选的轻量 NLP 流水线，基于 Phi-3 Mini 配合 LoRA 与 4-bit 量化，将文章分类为 8 个 PICO 标签。体积约 90 MB，可在消费级 GPU/CPU 上运行，加速生物医学研究中的证据综合。SLM-PICO-Screener is a lightweight NLP pipeline automating systematic review screening. Built on Phi-3 Mini with LoRA & 4-bit quantization, it classifies articles into 8 PICO labels. At ~90 MB, it runs on consumer GPUs/CPU, accelerating evidence synthesis in biomedical research.

系统综述与验证的质量模板——扩展 Consultant 角色原型。Quality template for systematic review and validation - Extends Consultant archetype

232M token 的 LLM 训练语料库：140 个颠覆性编程范式（60 个基础 + 80 个神经科学启发）+ 10 个神经 AI 表格数据集。真实神经科学内容：STDP、预测编码、自由能、海马体、Hebb。包含 JSONL+JSON+CSV。142,800 条范式条目 + 100K 神经样本。232M-token LLM training corpus: 140 disruptive programming paradigms (60 foundational + 80 neuroscience-grounded) + 10 neuro AI tabular datasets. Real neuroscience: STDP, predictive coding, free energy, hippocampus, Hebb. JSONL+JSON+CSV. 142,800 paradigm entries + 100K neuro samples.

60 个神经科学启发的软件架构范式。真实研究（Hebb、Bi & Poo、Friston、Buzsaki、Moser、Hodgkin-Huxley）映射到软件模式。61,200 条 JSONL 条目，约 82M token。全英文。60 neuroscience-grounded paradigms for software architecture. Real research (Hebb, Bi & Poo, Friston, Buzsaki, Moser, Hodgkin-Huxley) mapped to software patterns. 61,200 JSONL entries, ~82M tokens. All in English.

本地优先的印尼语字幕生成，面向授权学习资源。Local-first Indonesian caption generation for authorized learning sources

用于研究问题诊断和工作流支持的 Agentic 科学推理系统。Agentic scientific reasoning system for research-question diagnosis and workflow support.

HMS Endeavour Liber Primus 工作的公开证据、可复现研究与验证层。Public evidence, reproducible research, and verification layer for HMS Endeavour Liber Primus work.

珊瑚损伤后再生：PRISMA 系统综述与 meta 分析Coral regeneration after wounding: PRISMA systematic review and meta-analysis

珊瑚损伤后再生的 PRISMA 系统综述与 meta 分析。Coral regeneration after wounding: PRISMA systematic review and meta-analysis

Ultimate Rulesync AI Agent CLI Tool 2026 – Streamline Coding Workflows FastUltimate Rulesync AI Agent CLI Tool 2026 – Streamline Coding Workflows Fast

OncoRAG 是一个 RAG 系统，为临床医生和研究人员提供即时获取最新肿瘤学证据的能力——来源包括直接来自 PubMed 的随机对照试验、Meta 分析、系统综述和 III 期临床试验。提出临床问题，即可获得带引文的可靠答案。OncoRAG is a Retrieval-Augmented Generation system that gives clinicians and researchers instant access to the latest oncology evidence — drawing from Randomized Controlled Trials, Meta-Analyses, Systematic Reviews, and Phase III Clinical Trials sourced directly from PubMed. Ask a clinical question, get a grounded answer with citations.

每次工具调用，皆可验证。——AI Agent 工具调用的验证哨兵。捕获伪造、追踪成本、证明实际发生的事件。Every tool call, proven. — Verification sentinel for AI agent tool usage. Catches fabrication, tracks costs, proves what actually happened.

基于内容相似度的机器学习论文推荐系统，使用 Python 与自然语言处理技术。Machine learning recommendation system that suggests relevant research papers based on content similarity using Python and natural language processing techniques.

人工智能（AI）已成为提升乳腺 X 线摄影对乳腺癌诊断准确性的潜力工具。本系统综述与荟萃分析旨在探讨 AI 与放射科医生通过乳腺 X 线摄影检测乳腺癌的准确性。Artificial intelligence (AI) has emerged as a promising tool to improve the diagnostic accuracy of 49 mammography for breast carcinoma detection. This systematic review and meta-analysis aim to 50 explore the accuracy of AI and radiologists in breast carcinoma detection via mammography

面向 LaTeX 的实时引用管理与追踪工具，完全离线运行。Real-time citation manager and tracker for LaTeX. Fully offline.

后端实现 Agent Framework，用于编排 agent 工作流和消息交换。作为基于 Vercel 的客户端与 Microsoft 系统之间的内容和响应包装器，提供结构化的请求/响应处理、路由和生命周期管理，但不嵌入 AI 逻辑本身。Backend implementing the Agent Framework for orchestrating agent workflows and message exchange. Acts as a content and response wrapper between Vercel-based clients and Microsoft systems, providing structured request/response handling, routing, and lifecycle management without embedding AI logic itself.

自动化监控、打补丁与 Telegram 推送 Augment VSCode 插件更新，附带版本追踪与清理Automate monitoring, patching, and Telegram delivery of Augment VSCode plugin updates with version tracking and cleanup.

research questions、evidence maps、implementations 及可复现的工程 case studies。Research questions, evidence maps, implementations, and reproducible engineering case studies.

可复用的 Python 与 notebook 工作流，用于多语言科研数据准备、统计分析及 AI 文本检测器评估。Reusable Python and notebook workflows for multilingual research-data preparation, statistical analysis, and AI-text detector evaluation.

面向文档溯源 QA Agent 的生产级 Harness：检索、起草、自评、重写、升级循环，配套 LLM-as-judge 评测套件、分级权限工具、对破坏性操作的人工审批，以及一键容器化部署。A production-grade harness for a document-grounded QA agent: a retrieve, draft, self-score, re-draft, escalate loop with an LLM-as-judge eval suite, permission-tiered tools, human-in-the-loop approval for destructive actions, and a one-command container deploy.

面向 agentic AI 创业领域的战略情报平台——FastAPI + SQLite + 原生 JS SPA，内置专业 LLM agent 集群、咨询级行业档案、业务线领域矩阵以及 RAG copilot。Strategic intelligence platform for the agentic-AI startup landscape — FastAPI + SQLite + vanilla-JS SPA with a specialist LLM agent fleet, consulting-grade industry dossiers, a line-of-business domain matrix, and a RAG copilot.

个人 AI 工程博客与作品集——12 篇文章，覆盖 agentic AI、RAG、embeddings、chunking、token 及生产实践。基于 Astro 构建，部署于 Cloudflare Pages。Personal AI engineering blog and portfolio — 12 posts covering agentic AI, RAG, embeddings, chunking, tokens, and production patterns. Built with Astro and deployed on Cloudflare Pages.

面向 DeepSeek Harness agent 的学术写作工具包：论文大纲、标题/摘要模板、GB/T 7714 / APA / MLA 引用、措辞质检以及投稿前清单。Academic writing toolkit for DeepSeek Harness agents: paper outlines, title/abstract skeletons, GB/T 7714 / APA / MLA citations, phrasing QA, and a pre-submission checklist

26 个开源 Agent Plugins 1.0 包，统一 CLI 支持 Codex、ChatGPT、Cursor、GitHub Copilot、VS Code 和 Kiro。Skills + MCP，客户端定向交付，Apache 2.0。26 open-source Agent Plugins 1.0 packages with one CLI for Codex, ChatGPT, Cursor, GitHub Copilot, VS Code and Kiro. Skills + MCP, client-specific delivery, Apache 2.0.

🩺 为嵌入式医疗设备提供安全、合规级别的 AI 本地化方案，具备术语库控制与屏幕适配校验能力。🩺 Deliver safe, compliance-grade AI localization for embedded medical devices with glossary control and screen-fit validation.

📊 基于 Claude Skills 利用 XML 操作轻松创建和编辑 PowerPoint 演示文稿，实现精简的内容管理。📊 Create and edit PowerPoint presentations effortlessly with Claude Skills using XML manipulation for streamlined content management.

RepoMap AI 2026：为 AI 助手绘制突破窗口限制的代码库地图RepoMap AI 2026: Map Codebases for AI Assistants Beyond Window Limits

用于借助语言模型进行小说创作的全屏终端环境A full-screen terminal environment for writing fiction with language models.

学术论文推荐系统research paper recommendation system

学术论文推荐系统research paper recommendation sysytem

多工具编排 Agent，可针对复合型金融研究问题自动决策调用哪个金融数据 API，每项论断均附引用来源；包含配套的 Go 数据接入服务A multi-tool orchestration agent that answers compound financial research questions by deciding which of several financial-data APIs to call, grounding every claim in a cited source. Includes a companion Go ingestion service.