Havuz

BuildArena: Physics-grounded Besiege Building for LLMs

OpenCode agent configuration — a read-only orchestrator delegates to 6 specialized subagents. Expert model plans, fast models execute. Isola

MetAPI Go rewrite — meta-layer management and unified proxy for AI API aggregation platforms

Publication-quality LLM architecture figures from config.json — Raschka-style towers, diffs, posters, VRAM math. No weights needed.

The agent's prompt shows you the command; Termaxa shows you the consequence. Previews, backups, policy, an audit record. Claude Code, Codex,

The agent loop for Python. Your harness, any model.

Run the AEON Bench suite on your own hardware: verified HuggingFace pull → serve → benchmark (text · agentic ×3 harnesses · vision · audio ·

Strip AI writing tells at two layers: surface (Wikipedia Signs of AI writing) + narrative (StoryScope, arXiv:2604.03136). A model-agnostic L

CPU-first, eager-execution tensor/autograd runtime and LLM inference engine written in pure Zig

A juicy token compression algorithm. Compress upto 95% of tokens without any loss. Free & Opensource

Open-source JARVIS-class desktop AI. Zen mode orb, computer control, plugin marketplace, custom tools, local RAG, knowledge graph, voice per

MCP server for Homebox (>=0.26): inventory Q&A, intake, attachments, labels

The agent harness for Ruby applications. Durable sessions, streaming, human-in-the-loop approval, compaction, structured output, sub-agents,

Attention is your scarcest resource. Chief is the local-first layer that guards it — turning every agent, alert, and feed into one honest ca

Making MoE fine-tuning accessible to anyone with consumer hardware.

A modern terminal reimagining of the 1983 WarGames WOPR system — Cold War simulation TUI with AI-powered scenarios, DEFCON escalation, and m

Talk to your home. A voice-driven personal AI agent that chats, assists, and orchestrates your smart home.

Bilingual translation app and browser extension for Safari (iPhone/iPad/Mac), Chrome and Firefox — side-by-side web pages, dual subtitles, L

ANGLERFISH · Claude Fable 5 — light Binance-style quant dashboard (single HTML, live TradingView + 3D + Galton board) plus a safety-first Bi

Go LLM Execution Model — Go 原生 AI 编程 Agent CLI。单二进制、三层记忆、YAML 权限规则，TUI 对齐 Claude Code / Codex。

Graft, kod tabanından yerel bir code graph oluşturup coding agent'lara bağlamsal bilgi ve kalıcı proje kuralları sağlayan bir CLI aracıdır.

Neden yükseliyor · 86 günde 9.271 yıldızın yanı sıra son 28 günde 57 commit ve 10 farklı yazar, yüksek ilgiye eşlik eden kayda değer geliştirme hareketliliği gösteriyor. Agent'lara codebase context sağlama ihtiyacı yaygın; çoklu agent entegrasyonu da kullanım alanını genişletiyor. README'deki hız, token ve doğruluk kazanımları proje tarafından bildiriliyor; bağımsız doğrulama olarak değerlendirilmemeli.

OpenScience, bilimsel araştırma görevlerini planlayıp literatür taraması, kod çalıştırma ve sonuç raporlamayı tek bir AI workbench içinde birleştirir.

Neden yükseliyor · 86 günlük bir repoda 3.645 yıldız ve son 28 günde 434 commit dikkat çekici; bilimsel araştırma agent'ı fikri de güncel AI ilgisiyle örtüşüyor. README'de Y Combinator desteği belirtiliyor. Ancak aktivite yalnızca 5 farklı yazardan geldiği için büyümenin geniş katkıcı topluluğuna dayandığı söylenemez.

An open-source AI companion that actually remembers you — runs entirely on your own Mac or server. Desktop pet · long-term memory · proactiv

Free, open-source Grammarly alternative for Chrome & Safari — AI grammar checker and text rewriter for any website, powered by your own free

本地优先 AI 语音输入工具 | Local-first AI voice typing tool. 支持 Whisper/SenseVoice/Parakeet/Qwen asr，离线可用，LLM 智能润色. Open-source speech-to-text with LL

Make Claude Opus, Sonnet, GPT, or Gemini produce work close to what Claude Fable 5 would ship. 21 craft standards written and audited by the

A lightweight dashboard and reporting tool powered by OmniRoute - The Free AI Gateway

Run Claude Code workflow scripts, unmodified, on the OpenAI Codex CLI — fable plans, codex executes, fable verifies. Parallel agent fleets,

One GUI for native Claude Code, Codex, Cursor and more agents — manage them in one place, connect across devices. 一个 GUI，接入原生 Claude Code、Co

A compounding knowledge base maintained by LLM agents — inspired by Karpathy's LLM Wiki pattern

Cross platform GLM proxy for AutoClaw's Zhipu AI models. Provides OpenAI-compatible Chat Completions and Anthropic Messages APIs for Claude

The networked cross-session for coding agents — @mention another team's AI so yours and theirs integrate directly across sessions, machines,

Reina 桌面 AI Agent 发行仓库（仅安装包） | Release distribution for the Reina desktop AI agent — installers only. Windows x64 & macOS arm64, in-app auto

Fast WebGPU runtime for 1-bit (binary-weight) LLMs in the browser. Bit-exact, zero runtime dependencies.

A self-improving weather forecasting & trading agent, in public.

The first LLM-maintained knowledge wiki purpose-built for Product Managers — Karpathy's LLM Wiki pattern turned into a PM schema, rituals, a

A lossless token-optimization codec for LLM agents — cut input tokens up to 82%, nothing dropped, every fact kept. Drop-in proxy + one-line

A 100% local AI memory engine for your computer

Chrome extension for GEO workflow: capture LLM answers, analyze platform patterns, generate publish-ready articles — all local-first, open

A terminal coding agent written in Raven, with tools, subagents, skills, MCP, and multiple LLM providers.

Control plane for AI coding agents: route tasks, reduce token spend, run multi-agent workflows, fallback executors, and track cost per task.

DB증권 공식 Open API 파이썬 샘플코드 & SDK

Open-source, provider-agnostic AI browser agent (MV3 Chrome extension). Bring your own LLM: OpenRouter, OpenAI, Claude, Gemini, Groq, or a l

Self-hosted AI gateway — one endpoint, many providers, auto-fallback. OpenAI-compatible REST API with Cloudflare Workers AI automation.

Headless, deterministic audio engine for AI agents: compose scores as data, render byte-reproducible PCM offline, then listen through featur

Hermex, self-hosted Hermes agent sunucusunu iPhone’dan yönetmek için geliştirilmiş native SwiftUI uygulamasıdır.

Neden yükseliyor · 87 günlük yaşına göre 1.401 yıldız ve son 28 günde 237 commit dikkat çekici ilgi ve geliştirme temposuna işaret ediyor. Self-hosted agent’ları mobil cihazdan kontrol etme ihtiyacı da projeyi görünür kılıyor; ancak yıldızlar benimsemenin doğrudan kanıtı değil ve son dönemdeki commit’lerin yalnızca 2 farklı yazardan gelmesi geliştirme tabanının dar olduğunu gösteriyor.

Talos ağına bağlanıp yerel Ollama üzerinden açık model inference işleri çalıştıran bir GPU worker client.

Neden yükseliyor · 87 günde 686 yıldız almış; bu, kısa sürede görünürlük kazandığını gösteriyor. Ancak son 28 günde commit ve farklı yazar sayısı sıfır, dolayısıyla güncel geliştirme veya topluluk katkısı açısından destekleyici kanıt yok.

帮你的 Claude Science 一键接入你自己的 API：DeepSeek / 通义千问 / 智谱 GLM / Kimi / MiniMax / 小米 MiMo / 硅基流动 / OpenRouter / 任意 OpenAI·Anthropic 兼容端点

a community oriented 1:1, vLLM-alike (Continuous batching, paged KV) engine in C++ with additional features (GGUF, RadixAttention, Cache-awa

The black box for AI-written code. Passive decision graph linking every commit to the agent prompts behind it: drift detection, DLP audit of