仓库/Skill 库

本地优先、可审计的技术文档与源代码知识编译器。Local-first, auditable knowledge compiler for technical docs and source code

Claude Code 插件，用于科学论文写作、来源可追溯的审稿与去 AI 化润色。在统一类型化标准下汇聚 8 项 skill + 24 个工具：不可豁免的完整性阻断项、零目标 L0 词库、分级告警以及显式测量状态，避免缺失校准被误读为通过。针对 ApJ/MNRAS/PRD/JCAP 级期刊稿件进行了语料校准。Claude Code plugin for scientific paper writing, source-traced review, and de-AI rewriting. 8 skills + 24 tools under one typed standard: non-waivable integrity blockers, zero-target L0 lexicon, ranked advisories, and explicit measurement states so missing calibration never reads as clean. Corpus-calibrated for ApJ/MNRAS/PRD/JCAP-class manuscripts.

AI 原生时序记忆图，采用统一 DSL——将概念、查询、智能体工作流与语音统一为可部署的原语AI-native time-series memory graph with a single DSL — unifies concepts, queries, agent workflows, and voice into deployable primitives.

基于大语言模型的智能研发效能助手。深度解析需求文档，精准识别逻辑漏洞；多维度生成测试用例，提升用例设计效率；构建可复用的测试知识库，将个人经验转化为团队资产。赋能 QA ，实现质量保障智能化。

基于 Rust 构建的 MCP 原生嵌入式 AI agent 记忆数据库，提供 REMEMBER/RECALL/FORGET/SHARE 原语，支持混合向量搜索、AES-256-GCM 加密、DuckDB/PostgreSQL 后端以及 Python、TypeScript 与 Go 的 SDKMCP-native embedded memory database for AI agents built in Rust. REMEMBER/RECALL/FORGET/SHARE primitives with hybrid vector search, AES-256-GCM encryption, DuckDB/PostgreSQL backends & SDKs for Python, TypeScript and Go.

Proof Gradient 是 agent 演化协议：每次运行都会留下 proof，每个 proof 都会筛选 intelligence，每个被选中的 artifact 都会推动网络演化。Proof Gradient is the agent evolution protocol where every run leaves proof, every proof selects intelligence, and every selected artifact evolves the network.

Claude Code 扩展、hooks 和插件的实时索引，每 15 分钟从 GitHub 刷新一次。Live index of Claude Code extensions, hooks, and plugins — refreshed every 15 minutes from GitHub

面向 AI 编码 Agent 的跨工具、跨机器共享记忆。本地优先 SQLite，可选 Zenoh + RocksDB mesh，原生支持 MCP。Shared memory for AI coding agents, across tools and machines. Local-first SQLite, optional Zenoh+RocksDB mesh, MCP-native.

免费、本地优先、按章节编排的 LLM、向量数据库、RAG 与 Agent 课程——从零到你的首个 AI agent。Free, local-first, chapter-wise curriculum for LLMs, Vector DBs, RAG & Agents — from zero to your first AI agent.

Local-first、确定性的验证层，用于 AI 声明、记忆写入和风险操作。Local-first, deterministic verification layer for AI claims, memory writes, and risky actions.

无状态统一多提供商 AI 运行时，兼容 OpenAI 与 Anthropic 的 API，支持 Vercel UI 流式输出，覆盖 130k+ 模型的托管 Agent 与 Skills。Stateless unified multi-provider AI runtime with OpenAI & Anthropic-compatible APIs, Vercel UI streaming, managed agents and Skills across 130k+ models.

并行运行 Claude Code、Codex 与 Gemini，并支持彼此交接任务。面向 AI CLI 的便携 Windows 终端。Run Claude Code, Codex and Gemini side by side — and let them hand work to each other. Portable Windows terminal for AI CLIs.

Galactus 在低于规格的 Mac 上运行 744B 与 235B MoE 模型，具备与 llama.cpp 的逐位精度一致、以 RAM 作为缓存的执行模式，并提供完整的本地应用，涵盖 Agent、权限门控、代码编辑器、已认证的服务器模式、定时无人值守运行以及完整公开的测评数据Galactus executes 744B and 235B MoE models on undersized Macs with bit‑perfect llama.cpp parity, RAM-as-cache execution, and a full local app featuring an agent, permission gate, code editor, authenticated server mode, scheduled unattended runs, and fully published measurements.

[COLMW'26] AdaFlash：通过 On-Policy 蒸馏扩散 Drafters 实现的自适应投机解码[COLMW'26] AdaFlash: Adaptive Speculative Decoding via On-Policy Distilled Diffusion Drafters

多模型协作与智能辩论平台 — Multi-Agent 并行工作流 + RAG + 知识图谱，让多个 AI 像专家团队一样辩论、推理、共创

🔬 基于 AI 与研究论文对话，通过高级语义搜索与 RAG（检索增强生成）技术提取洞见与摘要。🔬 Chat with research papers using AI, extracting insights and summaries through advanced semantic search and Retrieval-Augmented Generation techniques.

DSH 完整移植版 DietrichGebert/ponytail — 懒惰 senior 模式，6 个中文 Skill，无空 tool

Music Assistant 的 FastMCP server——可从 Claude Code、Codex、Cursor 及其他 AI agent 控制 MA。FastMCP server for Music Assistant — control MA from Claude Code, Codex, Cursor, and other AI agents

系统综述与论文清单：探索面向代码的 LLM 中数据与代码质量问题的映射、检测与治理。Systematic review and paper list exploring the mapping, detection, and governance of data and code quality issues in Large Language Models for Code.

基于 Zotero 的文献综述助手，支持基于原文的 PDF 抽取、声明验证、内容综合及 Semantic Scholar 检索Zotero-based literature review assistant for grounded PDF extraction, claim verification, synthesis, and Semantic Scholar search.

利用 LLM 提示词、证据链接笔记、Excel 跟踪与跨论文综合，从 PDF 自动生成博士文献综述Automate PhD literature reviews from PDFs with LLM prompts, evidence-linked notes, Excel tracking, and cross-paper synthesis.

后端无关的 C# 库，用于构建有状态 AI 实体，具备按角色记忆、工具调用、情绪系统和自主 Agent 行为。Backend-agnostic C# library for building stateful AI entities with per-persona memory, tool access, mood systems, and autonomous agent behavior.

自动化并规模化 "LLMs as a participant"，将 LLM 作为研究参与者Automates and scales "LLMs as a participant."

自托管 AI 知识 Agent —— TypeScript、Bun、MCP。掌控你的数据，运行你自己的大脑。Self-hosted AI knowledge agent — TypeScript, Bun, MCP. Own your data, run your own brain.

Dust：一款受 React/Vue 启发的现代化、组件化 Dart Web 框架，基于 WebAssembly（WASM）构建，具有高性能与流畅的开发体验，可使用 Dart 构建健壮、类型安全的 Web 应用。Dust: A modern, component-based Dart web framework inspired by React/Vue. Built with WebAssembly (WASM) for high performance and a seamless developer experience. Create robust, type-safe web applications with the power of Dart.

面向 Model Context Protocol 服务器的 OWASP MCP Top 10 安全扫描器OWASP MCP Top 10 security scanner for Model Context Protocol servers

可自托管且功能丰富的视频库Self-hostable and feature-rich video library

为你的 AI 编程助手打造的持久记忆。它会记住你告诉它的内容，并在恰当时机把对应的笔记交还。本地机器全量运行。A memory for your AI coding assistant. It keeps what you tell it and hands the right note back at the moment it matters. Runs entirely on your own machine.

大语言模型元数据的开放注册表——以单一机器可读的 models.json 提供身份、作者、模态、上下文/输出限制、能力与生命周期日期，并通过 JSON Schema 校验。CC BY 4.0Open registry of large-language-model metadata — identity, authorship, modalities, context/output limits, capabilities & lifecycle dates as one machine-readable models.json validated by JSON Schema. CC BY 4.0.

面向多 Agent 工作流的 MCP 协调原语。LLM 无关、CLI 无关、可自托管。TypeScript + SQLite。MCP coordination primitive for multi-agent workflows. LLM-agnostic, CLI-agnostic, self-hostable. TypeScript + SQLite.

确定性的 MCP 安全架构。以 FrozenNamespace 作为 Model Context Protocol 工具验证的可信根（Root of Trust）。Deterministic MCP Security Architecture. FrozenNamespace as Root of Trust for Model Context Protocol tool verification

支持从多源智能检索、筛选与总结科学论文，提升研究与报告生成效率Enable intelligent retrieval, filtering, and summarization of scientific papers from multiple sources for efficient research and report generation.

🐙 通过 Python 与 PyTorch 实战代码深入理解 LLM 内部机制，掌握 LLM 架构、训练及核心原理。🐙 Explore the inner workings of large language models with hands-on Python and PyTorch code. Master LLM architecture, training, and mechanisms for AI systems.

实时更新的向量数据库项目、集成和基准评测全景图——每……刷新。Live-updating landscape of vector database projects, integrations, and benchmarks — refreshed every

GitHub 上 MCP server 的自动更新索引，每 15 分钟刷新一次。Auto-updated index of MCP servers shipping on GitHub, refreshed every 15 minutes