Havuz

AI 课程笔记：langchain / langgraph / python / vibe coding

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

Type-aware, evidence-backed delivery contract for AI coding harnesses. Different agents in. One predictable human handoff out.

Stellar MCP server for AI agents: one authenticated gateway to official docs, live ecosystem data, community intel, and tested playbooks.

Libra: Efficient Resource Management for Agentic RL Post-Training

Your agents. Your workspace. Your control—from prompt to commit, desktop to mobile. Works with Codex, Claude Code, OpenCode, Pi, Grok, DeepS

Deterministic-first execution engine for agent workflows in Go: the LLM extracts at the edge, a deterministic state machine decides. Zero de

The open-source talent graph for humans and AI agents.

A React + TypeScript spatial canvas and node-board library for whiteboards, visual editors, and graph interfaces.

Poirot is a deep research agent kernel built for those who care about how agents are architected.

MIT-licensed TypeScript runtime for AI-native game characters and worlds, with durable state, reliable actions, memory, tools, and Unity, Go

让你的 AI 智能体们组成一个团队：在任意宿主里把任务派给本地订阅的 Claude Code / Codex / Grok / Kimi / Antigravity，并行、不打断对话、全走已有订阅。

An open JSON format and local toolchain for carrying an AI assisted project's working state between agents, tools and model providers.

O Specsfy ajuda você a transformar uma ideia em software testado sem espalhar requisitos, planos e tarefas por vários arquivos.

StealthBench is a benchmark for evaluating whether autonomous agents operate with the operational security (OPSEC) discipline that separates

Local-first control plane for long-running and multi-agent AI systems. Observe, steer, recover and orchestrate agents.

An in-panel AI assistant and an MCP server for Filament v5 panels, built on laravel/ai and laravel/mcp: one tool list for both, the panel's

Automated friction logging for agents.

The business intent layer for your codebase. Ground user stories in the code that implements them.

Build AI Teammates that answer from your own knowledge.

MCP-compatible context and memory system: paged code navigation for AI assistants, plus a recorded, verified theory of what is known about t

Stateful execution layer for browser agents.

Gradient descent, but the parameters are agents — a parallel, asynchronous framework for self-evolving agents (skills, prompts, harnesses).

agent-marketplace is a command-line tool that installs shared AI agent resources into the AI coding tools you use

CAP plugin to build agents

TrueForge, LLM çağrıları, araç kullanımı, oturum durumu ve kullanıcı onaylarını yöneten; UI ve API üzerinden kullanılabilen açık kaynak bir agent runtime sunuyor.

Neden yükseliyor · Agent runtime alanına kapsamlı, çalıştırılabilir bir ürün sunuyor; README'de SDK, UI, sandbox, MCP ve dağıtım seçenekleri aynı çözümün parçaları olarak gösteriliyor. 66 günde 5.982 yıldız ve son 28 günde 289 commit ile 17 farklı yazar, güçlü ilgi ve kayda değer geliştirme hareketliliğine işaret ediyor; bu ölçümler tek başına üretim olgunluğunu kanıtlamıyor.

hSUM: Stateful Understanding & Memory — local evidence for agents. Keep the source. Return the passage.

Open-source local agent daemon for Agensis workspaces

Visual testing for agents: uiverify CLI, Vitest/Playwright SDKs, and agent skills.

Diff your AI agent's behavior between two runs. See exactly which tool calls, args, costs and outputs changed when you swap models or edit p

Unified Agentic AI and Data Platform

An agent-native runtime for embedded devices, powered by PocketJS. Build installable, updatable JavaScript apps for people and agents.

Internal agent platform on Amazon Bedrock AgentCore: interactive Claude Code cloud workspaces + headless agent kernels behind one portal, wi

AI-native, source-available (PolyForm Perimeter) game-dev toolkit: bforge — deterministic headless-Blender asset forge for AI agents (138 op

Lightweight observability for multi-agent AI pipelines. Self-hosted, no cloud dependency — add two lines of Python and see real-time agent e

Snap a line before your agents build.

Declarative automation for repetitive steps in herdr

The agentic debugger for Claude Code, Codex CLI and Gemini CLI: a token usage, cost and context-window tracker that shows what the tokens AR

Agent skills for Revit add-in development

Ask for the symbol, not the file. Token-efficient MCP server giving AI coding agents symbol-level reads of TS/JS, Rust, Python, Java, Go,

Small Rust CLI that turns your local browser into a structured search API for Claude Code, Codex, Cursor, OpenClaw, and shell agents—no host

The job hunt, in one window. With your agent. An open-source, agent-native, local-first desktop workspace for job discovery, browsing, docum

A token-spend profiler and cost-regression gate for AI agents.

A control plane so parallel coding agents stop fighting over the same simulator.

Build & prompt Qwen3.x agents in their native dialect — extracted from qwen-code & Qwen-Agent

TERSE - a hierarchical state language for humans and AI agents.

OpenShell Kubernetes Operator

Kiro Crew, geliştirme görevlerini oturumlar arasında sürdüren ve yerel ya da uzak donanımda çalışabilen bir agent workspace sunuyor.

Neden yükseliyor · 73 günlük yaşına kıyasla 4.159 yıldız ve son 28 günde 3.477 commit ile 20 farklı yazar, dikkat çekici ilgi ve yoğun geliştirme faaliyetine işaret ediyor. Sürekli çalışma, kalıcı görev bağlamı ve entegrasyonları tek bir geliştirme workspace'inde birleştirmesi de agent tabanlı araçlara yönelik güncel ihtiyaca denk düşüyor.

Connect AI to your data. Your metrics, your rules.

Treg, agent'ların tek bir API ve token üzerinden katalogdaki veya ekiplerinin kaydettiği araçları çağırmasını sağlayan bir tool registry ve proxy sunuyor.

Neden yükseliyor · 74 günde 3509 yıldız ve son 28 günde 1234 commit ile dikkat çekici görünürlük ve geliştirme etkinliği var; 7 farklı yazar katkısı da hareketin yalnızca tek kişiden gelmediğini gösteriyor. Bu veriler benimseme veya kaliteyi tek başına kanıtlamaz; README'deki katalog kapsamı ve kullanım iddiaları bağımsız olarak doğrulanmış değil.