多脚本多配置统一管理与自动化工具 | 轻松管理大量脚本并存储多个用户配置、设计自动化任务流、监看脚本日志,大幅提高自动化代理效率与稳定性!
这是一个采用 Next.js 构建的高颜值、毛玻璃(Glassmorphism)风格个人博客系统。本项目自带完善的前端展示与独立的本地后台控制台,支持 Markdown 沉浸式写作、草稿管理以及便捷的图床配置。(新增对移动端适配))
Набор навыков для AI(ИИ)-агентов (Claude Code, Cursor, Codex и др.), помогающий охватить полный цикл разработки на платформе 1С:Предприятие 8.3. Даёт модели абстракции над XML-форматами и CLI конфигуратора, а также глаза и руки для тестирования через веб-клиент. AI skills for 1C:Enterprise 8.3.
A genuinely good terminal
Annotate terminal text, review documents and agent replies in Herdr, and send the feedback straight back to the agent.
The platform for continuously improving AI agents.
Turn simple scene ideas into restrained, physically believable cinematic AIGC prompts.
Open-source models for financial risk detection and fraud analytics
A high-performance Rust implementation of the mihomo (Clash Meta) proxy kernel.
A simulation benchmarking platform for evaluating generalist robot policies
Code World Model Official Repo
Android BYOK LLM client with multi-provider access, agentic workflows, and remote device control.
将Chunqiu Detector-Problem solution写进 README并对其维护修改跟进等,方便求助者查阅。 (内容来自探索发现并测试,并非摘录自检测源码)
Give AI agents eyes, ears, and verifiable results. Watch Skill turns video, audio and screen activity into searchable, timestamped evidence and proves work with deterministic contracts, not model opinion. DeepWatch is the agent workspace built on DeepSeek Harness. Python + npm, MCP, CLI, REST, Web.
New X on Android compatible piko builds
A shadcn registry of Remotion components for product demo videos. Install a text reveal, device frame, terminal session, logo sting or a whole scene with a command.
Reproducing and studying RL algorithms for LLM agents, including GRPO, GSPO, DAPO, OPD, Search-R1, ReTool, ALFWorld and beyond.
An ESP32-based open source quadruped robot pet framework for developing Boston Dynamics-style four-legged robots that are perfect for STEM, coding & robotics education, IoT robotics applications, AI-enhanced robotics application services, research, and DIY robotics kit development.
倪海厦外门弟子,自学中医,为往圣继绝学,主要以倪海厦倪师人纪天纪为主,专攻中医经方,后期还有李可、胡希恕等中医老前辈授课内容。内含倪师人纪中医系列(非视频语音转文字,皆是倪师自编中医教材课本讲义) 自学顺序:针灸,黄帝内经,神农本草经,伤寒论,金匮要略五部教学课本PDF(请配合B站视频学习,搜索倪海夏即可) 自学中医笔记正在手打中...
Yet another HEN enabler for jailbroken PS5
Command Code provider plugin for DeepSeek Harness (dsh). Adds Command Code model access, live model catalog, plan-aware model selection, reasoning effort, image input, web search, and multi-account support.
A curated directory of 390+ official MCP servers from the companies behind the products. No unofficial forks, no abandoned projects.
Composable mathematics tools for agents
JevBench by Benchmark Heaven: benchmark and leaderboard for AI decision models, measuring typed decision accuracy, calibration, latency and cost.
Conditional integer multiplication: kappa > 2^-15. Community proofs, exact certificates and reproducible audits; assumes the OpenAI #109 framework.
Software reconstruction of the 1998 Yamaha FS1R synthesizer
Self-hosted media manager for movies, TV, music, ebooks and audiobooks: search, download and organise, all in one Docker container.
Claude Code plugin to see, review and iterate on your agent's work without leaving the terminal: show images, play videos, and run visual QA with Playwright recordings and before/after comparisons inside the Claude Code terminal. Works in any terminal that supports the kitty graphics protocol, fully cross-platform on macOS, Linux and Windows.
Doulor Cloud —— 一站式云端资源分发平台:二级域名 / 域名邮箱 / R2 直链网盘 / 个人名片 / AI API 中转站。纯静态前端 + Cloudflare Workers + D1 + R2。
Open Source Alternative to Durable Objects. Build multiplayer AI + complex agent coordination systems.
ComfyUI 姿势与机位编辑器 for Qwen Image 2.1:VNCCS 真人人偶 + OpenPose 一键摆姿(内置 209 骨架)+ 自由视角多人机位 | Pose & camera studio with VNCCS PoseStudio LoRA
—
Use your CodeBuddy subscription as local OpenAI APIs.
—
百度网盘直链解析助手 Flutter 客户端(Android / Windows):远端配置 + 直链解析 + 内置 Gopeed 多线程下载
Port games to work with the Steam Frame
Native PC port of Peter Jackson's King Kong (Xbox 360) via static recompilation
An open-source, cross-platform ebook and PDF reader.
轻书架官方 Flutter 客户端
An initiative to reconstruct and port official Touhou Project games and its derivatives to every platform.
天翼云电脑保活签到任务;移动云电脑保活,阻止休眠;web远程访问
A cute, lightweight desktop companion for Linux that idles, reacts, sleeps, and hangs out while you work.
FRAME · 为 RICOH GR II 扩展色彩与成片表达。自定义滤镜、负片、HDF、胶片颗粒、机内装饰与个性化关机体验。
🛸 让 AI 放手干活,让你心里有数。Mac / Windows 上的 Codex 额度、用量与估算成本小管家。
A curated list of custom firmware for music gear (and more). Links only.
Play PS5 game audio on any Bluetooth headphones or earbuds. Payload for jailbroken PS5
Deterministic code-graph context engine for AI coding agents - PostgreSQL + Apache AGE + pgvector, MCP & REST, no LLM in the loop.
RSIGame: Autonomous agentic game development with recursive self-improvement
Windows XP の Explorer を Wine 上で Linux のデスクトップシェルとして動かす / Run Windows XP's shell as the Linux desktop on Wine
C++/CUDA single-GPU inference engine for Qwen3.8 (incl. Qwen3.8-Flash-Next with offloaded experts) on the RTX 5090; grew from NInfer
A decompilation of Bloodborne (CUSA03173 v1.09) with its own runtime: the game's code rewritten as readable C++, one verified function at a time. The game doesn't run on it yet; 21 functions verified. No game files included.
A model-agnostic GPU runtime for shipping models as verified programs.
一个基于DeepSeek-Harness的ComfyUI插件
Adaptive Block Sparse Attention for MiniMax-H3
安卓AI聊天端:开箱即聊,无损导入酒馆卡,缓存强省用量,Jev 决策,专属RP模式,防空回,主动信息,前端卡支持,高定制主题,QQbot,AI群聊,插件系统,强兼容中转站。手机移动端原生支持,求 Star 收藏⭐Android AI Chatbox: lossless SillyTavern card import, aggressive cache optimization, Jev decisions, multi-dimensional memory, anti-blank-reply, proactive messages, AI group chat, SillyTavern themes, plugin system, Native mobile.RikkaHub
WorldPlay2: Extending Real-Time Interactive World Models in Control and Horizon
CLI for Grok Bot agents and groups: create, update, message, inspect threads, and automate cleanup
套餐拼图 · 把每个人想吃的拼成一桌:真实麦当劳菜单、心愿组合、官方核价与按人分餐。
Edge-triggered system monitor panel for Windows
WeeLLM runs large diffusion models with as little as 4 GB of VRAM, without any quantization. It dynamically determines how many layers can fit within the available VRAM and streams the text encoder and transformer layers to the GPU layer by layer, enabling inference on hardware with limited VRAM. It supports both safetensors and GGUF models.