Projects

### Overview: Context Mode optimizes AI coding agents by reducing context window size, improving performance, and enhancing user experience across multiple platforms. ### Key Features: - **Context Window Optimization:** Reduces context window size by 98%. - **Sandboxed Tool Output:** Ensures tool outputs are sandboxed and do not impact context size. - **Persistent Session Memory:** Stores session memory to prevent context loss. - **Routing Across 17 Platforms:** Routes tool calls across 17 different platforms using MCP + hooks. ### Tech Stack: TypeScript ### Why It Matters: Context Mode significantly improves the efficiency and performance of AI coding agents by optimizing context window size, reducing unnecessary data consumption, and ensuring consistent performance across various platforms.

Overview: This repository is a curated collection of papers, articles, and blogs related to data science and machine learning in production. It serves as a valuable resource for understanding how companies implement and optimize their machine learning projects. Key features: - Papers and articles on various machine learning techniques and their applications in production. - Summaries of ML advancements in tweets. - Real-world results and ROI assessments. - Guides and interviews on applying machine learning in different domains. Tech stack: None specified. Why it matters: This repository provides a comprehensive overview of the latest trends and practices in machine learning, making it an essential resource for developers and data scientists looking to stay updated and informed. Repository: eugeneyan/applied-ml GitHub stars: 30,619 GitHub description: 📚 Papers & tech blogs by companies sharing their work on data science & machine learning in production.

Overview: Composio is a powerful platform that powers 1000+ toolkits, provides per-user sessions, authentication, and a sandboxed workbench. It helps developers build AI agents that turn intent into action. Key features: - Composio SDKs for TypeScript and Python - OpenAI Agents, Claude Agent SDK, Vercel AI SDK, LangChain, and more providers - Per-user sessions, authentication, and triggers - A sandboxed workbench for testing and development Tech stack: TypeScript and Python Why it matters: Composio simplifies the process of building AI agents by providing pre-authenticated toolkits, per-user sessions, and a robust sandbox environment. This makes it an ideal choice for developers looking to quickly and securely deploy AI-powered solutions.

### Overview: This list compiles a diverse collection of AI autonomous agents, categorized into open-source projects and closed-source projects. It serves as a comprehensive guide for developers looking to integrate AI agents into their applications. ### Key Features: - **Adala**: A framework for autonomous data labeling and processing. - **Codeium**: A code interpreter for AI applications. - **Devon**: A platform for AI agents. - **Vortic AI**: A company offering AI solutions. - **UFO**: A project for AI agents. - **GPT Swarm**: A project for AI agents. - **Eidolon**: A project for AI agents. - **NexusGPT**: A project for AI agents. - **Brain Soup**: A project for AI agents. - **L2MAC**: A project for AI agents. ### Tech Stack: Python, JavaScript, and other relevant technologies. ### Why It Matters: This list provides a curated selection of AI agents, making it easier for developers to find and integrate these tools into their projects. The diverse range of projects ensures that developers can find solutions tailored to their specific needs, whether they are building their own agents or integrating existing ones into their applications.

Overview: OpenCLI is a powerful tool that converts any website into a CLI interface, allowing AI agents and humans to interact with websites and browsers in a deterministic manner. It also enables the automation of browser actions and the creation of new adapters for various websites. Key features: - Converts websites into CLI interfaces - Navigates, fills forms, clicks, extracts, and automates browser actions - Built-in adapters for popular websites like Bilibili, Zhihu, Reddit, and more - Supports AI agents for navigating and interacting with any website - Enables writing new adapters using the `opencli browser` command Tech stack: JavaScript Why it matters: OpenCLI simplifies the process of automating web interactions, making it easier for AI agents to interact with websites and browsers in a consistent and predictable manner. This tool is particularly useful for creating a bridge between human and AI agents, enhancing the capabilities of AI agents by enabling them to interact with websites as if they were human users.

### Overview: NautilusTrader is an open-source, production-grade Rust-native engine for multi-asset, multi-venue trading systems. It features a deterministic event-driven architecture and Python as the control plane for strategy logic, configuration, and orchestration. ### Key Features: - **Deterministic Event-Driven Architecture**: Ensures consistent and predictable behavior across all components. - **Python Control Plane**: Flexible and powerful for system composition and strategy development. - **Rust Engine**: High-performance trading engine with deterministic execution. - **Multi-Asset, Multi-Venue Support**: Scalable and robust for complex trading systems. - **Open-Source**: Free to use, modify, and contribute to the project. ### Tech Stack: Rust ### Why It Matters: NautilusTrader leverages Rust's performance and safety features to deliver a robust and reliable trading engine. The separation of concerns between the control plane (Python) and the execution plane (Rust) ensures that the system can be easily composed and developed using Python's flexibility while maintaining the performance and safety of a compiled engine. Repository: nautechsystems/nautilus_trader GitHub Stars: 30265 GitHub Description: Production-grade Rust-native trading engine with deterministic event-driven architecture.

Overview: The Stanford Alpaca project aims to build and share an instruction-following LLaMA model. The current model is fine-tuned from a 7B LLaMA model on 52K instruction-following data generated by Self-Instruct techniques. It is under development with limitations that need to be addressed. Key features: - Data generation and fine-tuning process - Code for generating the data - Code for fine-tuning the model - Recovering Alpaca-7B weights from weight diff Tech stack: Python Why it matters: The project aims to provide a reproducible model for research and evaluation, focusing on instruction-following capabilities and ethical considerations. Repository: tatsu-lab/stanford_alpaca GitHub stars: 30,233

### Overview: This repository is an open-source Data Science repository designed to provide a concise path for beginners to learn and apply concepts in Data Science. It aims to cover various aspects of the field, including algorithms, visualization tools, and training resources, to help users understand and apply Data Science principles in real-world problems. ### Key Features: - Tutorials - Free Courses - Massive Open Online Courses (MOOCs) - Intensive Programs - Colleges - The Data Science Toolbox: Algorithms, Machine Learning Packages, Deep Learning Packages, Visualization Tools, Miscellaneous Tools - Literature and Media: Books, Journals, Newsletters, Presentations, Podcasts, YouTube Videos & Channels, Facebook Accounts, Twitter Accounts, Telegram Channels, Slack Communities, GitHub Groups, Data Science Competitions ### Tech Stack: - Python, R, SQL, Data Visualization Tools ### Why It Matters: This repository serves as a comprehensive guide for anyone interested in Data Science, providing a structured approach to learning and applying the concepts. It covers the foundational knowledge needed to start a Data Science journey, making it an essential resource for beginners and professionals alike.

Overview: This repository hosts a collection of plugins designed to enhance Claude Cowork, a knowledge worker platform, by customizing its capabilities for specific roles and workflows. These plugins integrate seamlessly with Claude's core functionalities, allowing for more tailored and efficient solutions. Key features: - **Productivity**: Manages tasks, calendars, and workflows. - **Sales**: Handles prospect research, call preparation, and customer support. - **Customer Support**: Triage tickets, draft responses, and build knowledge base articles. - **Product Management**: Writes specifications, plans roadmaps, and keeps stakeholders updated. - **Marketing**: Drafts content, plans campaigns, and reports on performance. - **Legal**: Reviews contracts, manages NDAs, and prepares for meetings. - **Finance**: Prepares journal entries, reconciles accounts, and generates financial statements. - **Data**: Queries, visualizes, and interprets datasets. - **Enterprise Search**: Finds information across email, chat, and documents. - **Bio-Research**: Connects to preclinical research tools and databases. Tech stack: Python Why it matters: These plugins provide a flexible and customizable solution for knowledge workers, allowing them to leverage Claude Cowork's capabilities in a way that best suits their specific needs and workflows.

Overview: The OpenAI Agents SDK is a lightweight framework for building multi-agent workflows that supports various LLMs, including OpenAI Responses and Chat Completions APIs. Key features: - Agents: Configured LLMs with instructions, tools, guardrails, and handoffs. - Sandbox agents: Preconfigured agents for long-term work with a container. - Realtime agents: Powerful voice agents with `gpt-realtime-2.1` and full agent features. - Voice agents: Combines speech-to-text, agent workflow, and text-to-speech. - Agents as tools: Delegating to other agents for specific tasks. - Tools: Various Tools for taking actions. - Guardrails: Configurable safety checks for input and output validation. - Human in the loop: Built-in mechanisms for involving humans across agent runs. - Sessions: Automatic conversation history management. - Tracing: Built-in tracking of agent runs for debugging and optimization. Tech stack: Python. Why it matters: The OpenAI Agents SDK simplifies the process of building multi-agent workflows, making it easier to integrate with various LLMs and configure agents for different use cases. Repository: openai/openai-agents-python Primary language: Python GitHub stars: 30,104

Overview: EasyOCR is a ready-to-use Optical Character Recognition (OCR) library that supports over 80 languages and writing scripts, including Latin, Chinese, Arabic, Devanagari, Cyrillic, and more. It is compatible with Hugging Face Spaces and can be installed via pip or directly from GitHub. Key features: - Supports over 80 languages and writing scripts - Integrates with Hugging Face Spaces - Ready-to-use library - Compatible with various languages and scripts Tech stack: Python Why it matters: EasyOCR simplifies OCR tasks by providing a user-friendly library that can recognize text in multiple languages and scripts, making it ideal for developers and businesses needing to process text in diverse languages efficiently. Repository: JaidedAI/EasyOCR Primary language: Python GitHub stars: 30080 GitHub description: Ready-to-use OCR with 80+ supported languages and all popular writing scripts including Latin, Chinese, Arabic, Devanagari, Cyrillic, etc. README excerpt: # EasyOCR [](https://badge.fury.io/py/easyocr) [](https://github.com/JaidedAI/EasyOCR/blob/master/LICENSE) [](https

### Overview: Gitleaks is a powerful tool designed to help developers detect and manage sensitive information in their Git repositories. It uses regex-based rules to identify and alert on secrets like passwords, API keys, and other sensitive data. ### Key Features: - **Regex Detection**: Gitleaks leverages regular expressions to scan for patterns that match common secrets. - **File Scanning**: It can detect secrets in files, including those in `.gitignore` patterns. - **Commit Scanning**: Secrets can be found in commit messages and comments. - **Custom Rules**: Users can create and share custom rules for specific environments or projects. ### Tech Stack: - **Go**: The primary programming language used to develop Gitleaks. - **Regular Expressions**: The core technology for detecting secrets. ### Why It Matters: Gitleaks is essential for developers who need to ensure the security of their code and repositories. By automatically scanning for sensitive information, it helps prevent accidental exposure of credentials and other critical data. This tool is particularly useful in collaborative environments where multiple developers work on the same project. ### Repository: The Gitleaks repository is hosted on GitHub at [gitleaks/gitleaks](https://github.com/gitleaks/gitleaks). It has received over 30,000 stars, indicating its popularity and usefulness in the developer community. ### Getting Started: #### Installing: - **Homebrew**: `brew install gitleaks` - **Docker**: `docker pull zricethezav/gitleaks:latest` or `docker pull ghcr.io/g

### Overview: Modular Platform is an open-source platform for AI development and deployment, including MAX Framework and Mojo Language. It provides a unified environment for building, deploying, and managing machine learning models. ### Key features: - **MAX Framework**: A framework for building and deploying machine learning models. - **Mojo Language**: A programming language for building AI models. - **Mojo Compiler**: A compiler for the Mojo language. - **Mojo Standard Library**: A standard library for the Mojo language. - **Mojo Kernel**: A set of compute kernels for the Mojo language. - **Mojo Inference Server**: An endpoint for deploying models. - **Code Examples**: Examples of how to use the platform. ### Tech stack: Mojo ### Why it matters: Modular Platform provides a unified environment for AI development and deployment, making it easier to build, deploy, and manage machine learning models.

Overview: This repository contains 42+ runnable notebooks showcasing advanced techniques for Retrieval-Augmented Generation (RAG) systems, providing a comprehensive guide from foundational to cutting-edge methods. Key features: - Techniques for improving RAG systems - Detailed tutorials on various RAG techniques - Intuitive explanations, code implementations, and references for building accurate and context-rich retrieval systems Tech stack: Jupyter Notebook Why it matters: This repository offers a practical approach to understanding and implementing advanced RAG techniques, making it valuable for developers looking to enhance their retrieval systems.

Overview: This open-source book aims to make deep learning approachable, teaching concepts, context, and code through Jupyter notebooks. It is freely available, offers technical depth, includes runnable code, and supports community updates and discussions. Key features: - Interactive deep learning book - Multi-framework code - Math and discussions - Complemented by a forum for technical details and questions Tech stack: Python Why it matters: This book provides a comprehensive resource for learning deep learning, offering a balance between theory and practice, and fostering community engagement. Repository: d2l-ai/d2l-en Primary language: Python GitHub stars: 29,866

Overview: This repository serves as a comprehensive hub for generative AI research, courses, interview materials, and more. Organized around the question "what do you want to do?", it provides a variety of paths for users to explore and learn. Key features: - Courses and learning resources, by topic - Interview prep hub - Video transcripts and resources - Journeys for use, build, and understand AI Tech stack: HTML Why it matters: This repository offers a structured and organized approach to learning about generative AI, making it a valuable resource for anyone looking to understand or apply AI in their work.

### Overview: FastGPT 是一个基于 LLM 的知识型平台，提供数据处理、RAG检索、工作流编排等全面功能，简化了复杂应用的开发和部署。 ### Key features: - **Agent Skill 编排**: 简化工作流编排。 - **对话工作流、插件工作流**: 包含基础的 RPA 节点。 - **用户交互**: 实现双向 MCP。 - **双向 MCP**: 支持辅助生成工作流。 - **知识库复用**: 多库混用，支持手动输入和分段导入。 - **API 知识库**: 支持 TXT、MD、HTML、PDF、Docx、PPTX、CSV、XLSX 等多种文件格式。 - **RAG 模块热更新**: 实现热更新功能。 - **Agent-loop 热更新**: 实现热更新功能。 - **AI 实时生成插件**: 实现实时生成插件功能。 ### Tech stack: TypeScript ### Why it matters: FastGPT 是一个全面的 AI 工具平台，简化了知识型应用的开发和部署，使得开发者能够专注于业务逻辑，无需担心复杂的基础设施设置。

### Overview Infisical is the open-source platform for securing secrets, certificates, access, and AI Agents. It centralizes and manages sensitive information across teams and infrastructure, ensuring secure and compliant practices. ### Key Features - **Dashboard**: A user-friendly interface for managing secrets across projects and environments. - **Secret Syncs**: Seamless integration with popular platforms and tools like GitHub, Vercel, AWS, Terraform, Ansible, and more. - **Secret Versioning and Point-in-Time Recovery**: Track and recover secrets and project states. - **Secret Rotation**: Rotate secrets at regular intervals for services like PostgreSQL, MySQL, and AWS IAM. - **Dynamic Secrets**: Generate ephemeral secrets on-demand for services like PostgreSQL and MySQL. ### Tech Stack TypeScript ### Why It Matters Infisical simplifies the management of sensitive information for developers, making security more accessible and easier to implement. By centralizing and securing secrets, Infisical ensures compliance and reduces the risk of data leaks, making it a valuable tool for any organization looking to improve their security practices.

### Overview: Sim is a collaborative workspace designed for building, deploying, and monitoring AI agents and workflows. It is used by over 100,000 builders worldwide. ### Key features: - **Collaborative Workspace**: Teams can work together on AI projects. - **Integrated Documentation**: Accessible through the Sim.ai documentation. - **Slack Integration**: Direct communication through Slack. - **X Chat Support**: Support through the X chat platform. - **DeepWiki Integration**: Help and support through DeepWiki. - **Local Setup**: Easily set up on macOS using Docker Compose. ### Tech stack: TypeScript ### Why it matters: Sim is a powerful tool for developers and AI enthusiasts, offering a seamless environment for building, deploying, and managing AI agents and workflows. Its widespread use by over 100,000 builders underscores its importance in the AI development ecosystem.

Overview: This repository is a comprehensive collection of cybersecurity references, scripts, tools, code, labs, and training resources, curated and maintained by Omar Santos. Key features: - Comprehensive cybersecurity resources - Domain taxonomy for offensive and defensive security - AI security, LLM engineering, and ethics resources - Certifications for various domains - Build-your-own lab resources - Training reference for cheat sheets and curated projects Tech stack: Jupyter Notebook Why it matters: This repository serves as a valuable resource for cybersecurity professionals, providing a structured and curated collection of materials for offensive, defensive, and AI security, as well as training and certification preparation.