opensandbox-group/OpenSandbox
OpenSandbox is a general-purpose sandbox platform providing secure, extensible runtime environments for AI applications, supporting various language SDKs and Kubernetes for scalable deployment.
Awesome AI Agents › Weekly › 2026-06-06
2026-06-06
AI Agent Development
TencentCloud/TencentDB-Agent-Memory
TencentDB Agent Memory provides a 4-tier progressive pipeline for local, long-term AI agent memory, enhancing performance and reducing token usage without external API dependencies.
agentscope-ai/agentscope-java
AgentScope Java is an agent-oriented programming framework for building LLM-powered applications with features like ReAct reasoning, tool calling, memory management, and multi-agent collaboration.
AI Agent Orchestration
dromara/liteflow
LiteFlow is a lightweight, high-performance, and programmable Java-based component-based rule engine that seamlessly integrates AI Agents for complex business orchestration and hot deployment.
AI Agents
NVIDIA/NemoClaw
NVIDIA NemoClaw is an open-source reference stack for securely running always-on AI agents like OpenClaw and Hermes within NVIDIA OpenShell sandboxes, offering managed inference, network policy, an...
Agent Monitoring Tools
junhoyeo/tokscale
Tokscale is a CLI tool and visualization dashboard for tracking token usage and costs across various AI coding agents like OpenCode, Claude Code, and Gemini CLI.
Documentation & Testing Assistants
siyuan-note/siyuan
SiYuan is a privacy-first, self-hosted, fully open-source personal knowledge management system with rich features like block-level referencing, Markdown WYSIWYG editing, and AI integration, support...
Language Agent Frameworks
chopratejas/headroom
Headroom is a context compression layer for AI agents, significantly reducing token usage (60-95%) across various inputs like tool outputs, logs, RAG chunks, and conversation history, while maintai...
Multi-Agent Frameworks
strands-agents/harness-sdk
Strands Agents is a versatile SDK for building and operating AI agents with a model-driven approach, supporting multiple LLMs and multi-agent systems.
Prompt Libraries
Jeffallan/claude-skills
A collection of 66 specialized skills designed for Anthropic's Claude model, aiming to enhance its capabilities for full-stack development tasks and operate as an expert pair programmer.