Fosowl/agenticSeek
AgenticSeek is a fully local, autonomous AI assistant powered by Deepseek R1 agents that can code, browse the web, and manage files without relying on cloud services, ensuring user data privacy and eliminating ongoing costs.
Awesome AI Agents › Programming Languages
smartcomputer-ai/agent-os
Agent OS is an experimental framework and runtime designed to build sophisticated, long-running, and self-coding autonomous AI agents. The core idea behind Agent OS is to provide a computing substrate where AI agents can write and execute their own code to interact with the world effectively. This environment is tailored to support self-evolving agents that can operate robustly over extended periods. The project is inspired by the concept of next-generation Auto-GPTs and aims to create a local-first system that runs primarily on the user's machine, with optional cloud components. The current implementation is in Python, but the architecture is language-agnostic and is planned to be rewritten in a systems programming language for better performance and scalability. Agent OS integrates well with other AI libraries such as LangChain and LlamaIndex, enhancing its capabilities for building intelligent agents. A novel data layer inspired by Git's content-addressable storage ensures longevity and reliability by maintaining the agent's history and memory, preventing data loss. The primary function of an agent within this system is not just to run large language models locally but to execute generated code and functions, composing them into larger features and managing event and data routing. The project includes a demo agent called "Jetpack," which showcases the platform's capabilities by engaging in conversations and writing self-executing programs. Jetpack features a web interface hosted by the Agent OS web server, allowing users to interact with the agent easily. The repository contains various modules, including an object store, state transition function libraries, runtime components, synchronization tools, a web server, and a command-line interface. Extensive unit tests ensure code quality and reliability. Agent OS is currently in alpha and actively developing a distributed runtime to enhance its functionality. It supports multiple operating systems and requires Python 3.10 or higher and Poetry for dependency management. The project is well-documented with design documents, a manifesto explaining its motivation, and usage instructions to help users get started quickly.
https://github.com/smartcomputer-ai/agent-os
Also in Programming Languages
MODSetter/SurfSense
SurfSense is a highly customizable AI research agent that integrates with personal knowledge bases and external sources to provide advanced, privacy-conscious research and knowledge management capabilities.
HKUDS/AI-Researcher
AI-Researcher is a fully automated scientific discovery platform powered by large language models that autonomously conducts research from idea generation to manuscript creation.
InternLM/xtuner
XTuner is a comprehensive and efficient toolkit for fine-tuning large language and vision-language models, supporting a wide range of models and training algorithms with scalability and flexibility.
developersdigest/llm-answer-engine
A Perplexity-inspired answer engine built with Next.js and multiple AI and search technologies to provide rich, context-aware responses including sources, images, videos, and follow-up questions.
run-llama/LlamaIndexTS
LlamaIndexTS is a lightweight, extensible data framework designed to integrate large language models into server-side JavaScript applications, supporting multiple runtimes and a wide range of LLM providers.
EgoAlpha/prompt-in-context-learning
An open-source comprehensive resource and engineering guide for mastering in-context learning and prompt engineering with large language models like ChatGPT, GPT-3, and FlanT5.
AnotiaWang/deep-research-web-ui
Deep Research Web UI is a secure, AI-powered web interface that enables iterative, deep research by combining search engines, web scraping, and large language models with real-time feedback and exportable research reports.