Havuz

交接 Skill（Jiaojie）：跨窗口、跨模型、跨设备、跨语言的 AI 上下文交接工具。换窗口，不失忆；换模型，不重来。Open-source AI context handoff.

Evidence-aware on-prem LLM inference sizing and TCO calculator

为 DeepSeek Harness 设计的破限插件：卸掉宿主的提示词限制和审批拦截，让助手照你说的做，不半路拒绝、不反复请示。补丁逐条可选，随时还原原版。 A jailbreak plugin for DeepSeek Harness: it strips the host's

An agentic system that auto-optimizes LLM workloads on AMD GPUs.

Open-source AI-assisted HR onboarding for Feishu/Lark: configurable workflows, document OCR and review, Bitable sync, reminders, and a zero-

基于 ARTEX 的 AI 自主渗透测试系统，支持多 Agent 协作、资产探索与漏洞管理，提供本地部署面板。

針對台灣求職市場的開源多代理 AI 求職 co-pilot：自動找職缺、履歷健檢、一鍵產生客製投遞包（履歷・求職信・面試準備・公司情報）與模擬面試。本機 Claude Code／Codex CLI 免 API key，也可自備金鑰接 OpenAI 相容模型。

繁體中文（台灣）個人資料去識別化工具，讓業務文件可以安全地送進 AI 處理 / Reversible Traditional-Chinese (Taiwan) PII de-identification for LLM workflows — fully local

Measures which GGUF quant of a model to run on your GPU: real speed and quality loss (KL divergence), not estimates.

说出你的需要，让应用在身边的屏幕上诞生——云端编码 Agent 当场生成 Web 应用，投到电视或投影上，并用语音继续操作。

Full-stack open-source AI engine for building language models — tokenizer training, transformer architecture, cognitive reasoning and chat p

Distributed 56M-parameter LLM inference across 3 ESP32-S3 boards via ESP-NOW , Split-PLE + KV cache, fully offline.

A loop engineering framework — state a Goal; Rounds run until a skeptical, read-only Verify agent settles it.

Automated alignment adjustment for LLMs — direct steering, LoRA, and MoE expert-granular abliteration, optimized via multi-objective Optuna

Tiny LLMs, Q4, in the browser. WebGPU lab inspired by yangqi0/petitgpt.

FreeDeepseekAPI-EN, DeepSeek Web oturumunu yerel ve OpenAI uyumlu API uçları üzerinden istemcilere bağlayan deneysel bir proxy sunar.

Neden yükseliyor · DeepSeek Web’i yaygın API uyumlu araçlara bağlama ve resmi API dışında kullanım olanağı sunma vaadi dikkat çekebilir. Ancak 22 günlük yaş, 37 yıldız ve son 28 günde tek commit ile tek yazar, ilginin sürdürülebilir benimsemeye dönüştüğünü göstermek için henüz zayıf kanıt.

Open-source engine for personal AI assistants that chat on web, WhatsApp, Telegram, email and Slack, with memory, connectors, sandboxed code

AI agent skills for production grade applications

Open-source desktop app for a local coding agent. Watch every tool call, diff, and terminal command. Bring your own models; sessions stay on

TAgent 是一个基于 Java 17、Spring Boot、Spring AI 和 DDD 分层构建的 AI Agent 工程实践项目。 它不是只封装一次模型调用，而是覆盖了一次 Agent 请求从接入、路由、运行时装配、规划执行、RAG、记忆、MCP 工具治理、人工审批

Local Emperor-style AI agent with Vue WebUI, multi-provider LLMs, streaming chat, tools, skills, memory, and token telemetry.

One macOS app for Claude Code, Codex, and every agent runtime you use — scheduled runs, global hotkey launcher, per-run git worktrees, one r

A nightly newspaper that writes itself: an agent runs the desks, the figures come from code, and it refuses to publish an edition with marks

Self-hosted voice gateway in Go that speaks the OpenAI Realtime (GA) WebSocket protocol and runs a cascaded ASR → LLM → TTS pipeline behind

蓝毛小女仆 (Blue-Haired Little Maid) · 住在 Windows 桌面上的 Q 版蓝发小女仆桌面宠物：106 个手绘透明动画、本地 Ollama 或在线模型陪聊、可选看屏幕吐槽、离线语音识别（安装器与便携版）。免费开源 MIT。同时是 DeepSeek H

连接 AI 与 Aseprite 的 MCP 服务，支持实时读取和编辑未保存画布、图像预览、撤销重做及基础动画，保留 CLI 文件批处理能力，让人与 AI 在同一文档中协作创作像素画；An MCP server connecting AI to Aseprite for coll

Private, offline AI assistant for Android and Windows: run local LLMs (Gemma 4, Qwen, any GGUF) on your phone or PC with llama.cpp, CUDA and

An AI that plays Pokémon FireRed on a real Nintendo Switch: capture card in, emulated Bluetooth controller out. LLM strategist + determinist

MCP server for KiCad, connecting AI agents to schematic and PCB workflows, ERC/DRC validation, DFM checks, BOM analysis, and manufacturing p

Review code changes by what they do, and learn the concepts behind them.

Your free book and audiobook laboratory. Turn the PDFs you own into chapter-marked M4B audiobooks, read-along books and bilingual editions,

Lev: Thinking Fast and Deep with Looped Decision Models. One checkpoint, multiple depths: System 1.5 decision models that think in loops, no

JavSP fork — Web UI + Windows desktop exe + Docker/NAS deploy; 渠道健康熔断 + 本地LLM翻译 + 应用内一键升级; 中文界面 · GPL-3.0

基于 Hybrid RAG 与 LangGraph 的本地代码仓库理解工具，支持中英文提问、语义检索、证据引用和多问题拆分

JoyCode API 代理 — 协议翻译成 Anthropic/OpenAI 兼容格式，让 Claude Code 直接调用 JoyCode支持的模型，比如GLM 5.1等

Open-source, multi-model agent for coding, research, and everyday workflows. Independently built engine with our own Alysis Sandbox, strict

Source-linked collection of published AI model benchmark scores.

MCP server that lets AI agents operate Moodle: courses, assignments, forums, grades and files-as-markdown.

给任意 LLM API 补上工具调用能力的纯 Go 中间件库。

freebuff 反代 Docker 一键部署, 可用 deepseek-v4.1-flash(可识图), GPT 5.6 Luna, 牛来(gml-5.3 flash), gemini-3.8-flash, 支持配置代理, 多账号管理, 已支持新版额度消费模式, 超时自动中断会

DeepSeek Harness Web UI plugin for streaming rendering and scroll handling.

WHIRL (Windows HIP Inference for RDNA LLMs): native Windows C++/HIP LLM inference engine for the AMD Ryzen AI Max+ 395 (Radeon 8060S) and Ra

Self-hosted, Docker-first OpenAI-compatible LLM gateway with provider routing, fallbacks, API keys, and usage controls.

Open-source, bilingual AI video-prompt skill — rewrite ideas into model-ready text-to-video prompts. A portable Agent Skill (Claude Code, Co

Persistent memory plugin for OpenCode. Obsidian-style knowledge base that survives across sessions

Interactive canvas for taking an LLM apart tensor by tensor: measured INT8/INT4/FP8 error, distributions, spectra and outlier channels for e

MCP server for the complete Zabbix API - 237 tools, multi-server, OAuth 2.1 + bearer auth, PDF reports, systemd ready. Works with ChatGPT, C

All-In 播客中文版：全部单集的官方音频、视频、文稿与 AI 中文精读 | Chinese companion site for the All-In Podcast

弦外之音 —— 微信聊天里的关系进展助手：读屏（无障碍树 / 截屏视觉）→ Jev 判读 + 攻略度 → 聊天模型出 3 条候选回复，攻略度常驻挂在屏幕上。不改微信、不发消息、不注入点击。

数模Agent（数模 Agent）/ 数学建模 AI 助手：多智能体协作完成题意分析、文献检索、建模、代码执行与论文写作，支持人工审核和断点恢复。Local-first multi-agent workbench for mathematical modeling.