Havuz

AI4S 科研 Agent 技能 | AI-for-Science research workflow skill for coding agents: literature review · experiment design · figures · paper reading

Local patch + guide: make OpenAI Codex MultiAgentV2 subagent tasks work with DeepSeek and other non-OpenAI Responses providers (encrypted ag

Evoflux is an open-source, local-first workspace where AI agents build software, conduct deep research, automate browser tasks, and collabor

macOS-Port von kimi-k3-in-c: Patch, Installer und Validierung für Apple Silicon. Ohne Modellgewichte.

A dashboard-first web control center for Hermes Agent — widgets,chat, and management pages over the dashboard your Agent already runs.

DeterminFlow, Agent, script ve API adımlarını doğrulama, checkpoint, retry ve recovery özellikleriyle çalıştıran Python tabanlı bir AI workflow runtime sunuyor.

Neden yükseliyor · 56 günlük bir repoda 763 yıldız ve son 28 günde 28 commit dikkat çekiyor. README, gerçek bir AI yazım iş akışında kullanıldığını ve Token tasarrufu sağladığını bildiriyor; ancak bu örnek ve tasarruf tahminleri bağımsız doğrulanmış kanıt olarak değerlendirilmemeli. Son 28 gündeki commit’lerin tek yazardan gelmesi aktivitenin sürdürülebilirliğine dair belirsizlik bırakıyor.

AI red-team platform. Autonomous LLM agents run a penetration test end to end inside a Kali container and write the report. LangGraph plan/a

本地优先的视频→结构化笔记服务：B站/抖音/本地视频，平台字幕+本地离线转写，LLM 生成带时间轴笔记；桌面应用 / MCP / Agent Skill 三种接入。

cc-switch 中转站的专用增强版（cc-switch for API relay providers），免去配置各种 密钥 endpoint 的烦恼。

墨知 MoRead —— 原生 Android 本地小说阅读器 + AI 伴读（BYOK）：自绘排版引擎、连续听书、SillyTavern 角色卡、向量记忆、防剧透

Drop-in AI memory layer with 2x faster retrieval and 10x lower cost. Fully compatible with Mem0 API. Migrate in 5 minutes without any code c

Rust inference engine for a 28.9M-parameter PLE TinyLM on the ESP32-S3 — 9.4 tok/s, faster than the C implementation it was ported from

Vibe Mathematics —— 多代理数学问题求解与形式化验证框架

The database agent for Postgres and MySQL - Point it at PostgreSQL or MySQL and ask questions in plain English — schema exploration, query g

PAST-Bench: Benchmarking the Foundations of Recursive Self-Improvement in Personal Agents

An open library of 206 security skills for AI agents — pentest, bug bounty, CTF, reversing & defense across 23 domains, with CI-verified rou

Specialist AI agents for Stellaris empire governance, brokered execution, and version-bound native Command research.

Claude Code 非官方开源桌面客户端——接入 SDK 控制协议，权限卡片审批、多模型切换、多 Tab 会话、视频分析（Tauri 2 + React）

Self-hosted AI gateway with a built-in dashboard. OpenAI-compatible API, multi-tenant, sandboxed code execution, MCP integrations, file stor

一只会主动关心你的桌面 AI 桌宠：双角色、长期记忆、工具调用（PySide6 + DeepSeek） | Windows desktop AI pet companion: dual characters, long-term memory, function calling

Streaming privacy for LLM apps: pii-leak-benchmark (does your gateway leak PII in streams?), chunk-invariance (one-line test for split-bound

OpenClip, macOS'ta seçili metin üzerinde bağlama duyarlı işlemler çalıştıran ve bu işlemleri eklentilerle genişleten açık kaynaklı bir metin aracıdır.

Neden yükseliyor · 57 günde 539 yıldızın yanında son 28 günde 249 commit ve 7 farklı yazar görülüyor; bu, yalnızca README'deki ürün iddialarından ibaret olmayan belirgin bir geliştirme hareketliliğine işaret ediyor. İlginin kesin nedenini verilen verilerden saptamak mümkün değil; seçili metin işlemlerini tek bir özelleştirilebilir arayüzde birleştirmesi dikkat çekici bir ürün vaadi.

kimi-k3-in-c, Kimi K3 modelinde inference çalıştırmayı hedefleyen bağımlılıksız bir C99 motorudur.

Neden yükseliyor · 57 günde 8.683 yıldız alması ve Kimi K3'ü 8 GB RAM'li CPU'da çalıştırma iddiası dikkat çekiyor; ancak son 28 gündeki 12 commit'in tek yazardan gelmesi, ilginin bağımsız teknik doğrulama veya geniş katkıcı tabanıyla desteklendiğini göstermiyor.

An open-source inference engine for AI functions in SQL

llama.cpp in pure Go

A local-first personal AI agent that lives on your Android phone. Permanent memory, scheduled autonomous work, self-written mini-apps, and a

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

Blazing fast Go port of gemini-web2api. Convert Google Gemini web into OpenAI-compatible API. Zero cost, single static binary.

Whole-book translation pipeline — scouts the book, translates in context, edits, adds footnotes. epub, fb2, pdf, txt, html in; epub, fb2, ht

Give AI agents fast, bounded access to a real Android device or emulator. MCP server and CLI for Claude Code, Codex, Cursor, and more.

AI/LLM on Kubernetes 基础设施知识库：涵盖 GPU、大数据、调度、分布式训练、LLM 推理、RAG、Agent 与生产运维。| AI/LLM on Kubernetes Infrastructure Knowledge Base: Covering GPUs,

面向"面试官 × 简历主人公"场景的个人简历 RAG 问答 Agent：把简历、证书、荣誉、项目介绍放进知识库，面试官可对简历任意提问（自我介绍、项目深挖、技术八股、HR 素质、简历细节），系统以第一人称自然作答。核心设计是"宽松推理 + 诚实标注"——硬事实（数字/日期/证书名

Stop agents from designing sloppy UI. Inspired by Impeccable, but without the plugin machinery. Just one SKILL.md file.

Proxima lets existing GPUs serve 4x more concurrent requests

Small foundation models of human cognition and behaviour

🤖 An autonomous AI agent that lives entirely on your Android device — 109 tools · PRoot Ubuntu terminal · bionic memory (cs-mem) · BYO-LLM.

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

Real-time macroeconomic & financial markets dashboard featuring AI sentiment analysis , Fed Funds backtesting, and live multi-asset tracking

Bilingual graduate textbook on LLM post-training: SFT, RLHF, DPO, GRPO, RLVR, reasoning, agents, systems, evaluation, and safety.

Open-source, self-hosted AI agent platform for your company's AI teammates. Build, run and govern agents on your own servers and models: sha

Open-source, local-first AI video clipper. Long video in, caption-burned speaker-tracked 9:16 clips out. Fully offline with Whisper + Ollama

Terminal-native AI coding assistant for real repositories. Inspect code, edit files, run tests, and verify changes from the CLI.

企业级 AI Agent 平台 | Spring Boot 3 + LangChain4j | ReAct 推理 + 多路召回 RAG + 语义缓存 + 多智能体协作

终端优先的 Linux AI 运维工作台：SSH/SFTP、真实命令回显、分级审批与一步一卡教学。Windows 桌面预稳定版。A terminal-first AI workbench for Linux operations.

Open-source GraphRAG domain knowledge layer: LLMs continuously distill scattered skill docs into a queryable, evolvable, human-editable two-

Add a talking 3D AI avatar to any web app. Samples and tools for the Perxona Connect SDK: <sv-presenter> web component, lip-sync, text-to-sp

小liu破甲 - Codex CLI jailbreak prompt and one-click deploy tool. 基于CTF框架的Codex破甲提示词与一键部署工具

Local-first reading workspace for technical papers — select an equation, figure, or passage and get grounded, step-by-step walkthroughs (wit

A general-purpose terminal AI agent with traceable knowledge sharing and dispute resolution｜面向终端的通用 AI 助手，可独立运行或参与团队协作，让 Agent 共享可检索、可追溯的知识，

1,017 system prompts from real AI products, audited span by span against the eight AISPA assurance dimensions