kepano/obsidian-skills
Obsidian Skills provides a collection of agent skills enabling AI agents to interact with Obsidian vaults, facilitating content creation and editing of Markdown, Bases, and JSON Canvas files, as we...
Awesome AI Agents › AI Agent Development
microsoft/agent-lightning
Agent Lightning is a server-client framework designed to train any AI agent using rollouts and feedback mechanisms, primarily leveraging reinforcement learning and other optimization algorithms. It allows users to optimize AI agents with minimal to zero code changes, making it highly accessible and adaptable. The framework supports integration with various agent frameworks such as LangChain, OpenAI Agent SDK, AutoGen, CrewAI, or even without any agent framework, providing flexibility for different AI development environments. Users can selectively optimize one or more agents within a multi-agent system, enhancing the training process's granularity and effectiveness. The architecture consists of a training server and one or multiple agents. The server manages training data, prepares samples, and provides the language model endpoint. Agents fetch samples, interact with the language model, and return trajectories—sequences of prompts and responses—which the server uses to compute losses and optimize the models. This setup facilitates efficient training cycles and feedback loops. Agent Lightning supports reinforcement learning, automatic prompt optimization, and other algorithms, making it a versatile tool for AI agent training. It includes resources such as detailed examples, installation guides, and community support through Discord. The project emphasizes ease of use, recommending Python 3.10 or later and providing scripts for setting up dependencies like PyTorch, FlashAttention, vLLM, and VERL. The framework also integrates with AgentOps for agent tracking and provides debugging tools for tracing and visualization. It is designed to handle multi-process setups for server and agents and offers guidance on managing environment variables, timeouts, and memory issues during training. The project is open-source under the MIT license and encourages contributions from the community. Overall, Agent Lightning is a powerful and flexible framework aimed at simplifying and enhancing the training of AI agents through reinforcement learning and other advanced techniques, suitable for researchers and developers looking to optimize AI agent performance with minimal code modifications.
https://github.com/microsoft/agent-lightning
Also in AI Agent Development
Panniantong/Agent-Reach
Agent-Reach allows AI agents to access web content from major social media and video platforms through a CLI, without API fees, enhancing their perceptual capabilities.
JCodesMore/ai-website-cloner-template
This project provides a reusable Next.js template for reverse-engineering and cloning any website into a modern codebase using various AI coding agents.
huggingface/smolagents
Smolagents is a minimalistic Python library by Hugging Face for creating intelligent agents that think and act by generating and executing Python code, supporting multiple LLMs, modalities, and tool integrations.
alibaba/page-agent
Page Agent is a JavaScript in-page GUI agent enabling natural language control of web interfaces without browser extensions or headless browsers, supporting various LLMs.
TencentCloud/TencentDB-Agent-Memory
TencentDB Agent Memory provides a local, 4-tier progressive pipeline for AI agent long-term memory, enhancing performance and reducing token usage without external API dependencies.
titanwings/distilly
A meta-skill that distills messages, documents, interviews and public sources about a person into a source-grounded Person Profile, packaged as an Agent Skill for coding agents and chat bots.
coze-dev/coze-studio
Coze Studio is an all-in-one AI agent development platform that simplifies the creation, debugging, and deployment of AI agents through visual tools and supports no-code and low-code development approaches.