Awesome AI Agents › Inventory Automation Agents
Agent Workflow Memory (AWM) is a project that introduces a novel approach to inducing, integrating, and utilizing workflows through an agent memory system. A workflow in this context refers to a common sub-routine used in solving tasks, where specific example contexts are abstracted out to create reusable processes. The project supports both offline and online operational modes. In the offline mode, workflows are induced from ground-truth annotated examples when additional training data is available. In the online mode, workflows are induced dynamically from past experiences without relying on auxiliary data. AWM has been tested and demonstrated on two main platforms: WebArena and Mind2Web. On WebArena, AWM achieves a state-of-the-art success rate of 35.6%, showcasing its effectiveness in task completion within this environment. On Mind2Web, AWM performs exceptionally well among text-based agents, with the offline mode showing strong generalization capabilities across various tasks, websites, and domains. The project includes detailed instructions for running AWM on both platforms, with specific setup commands and environment configurations provided in the respective directories. The repository also features visual aids and performance charts that highlight the success and benchmarking results of AWM. The project is academically grounded, with a citation available for reference, indicating its research contribution to the field of agent-based workflow memory systems. Overall, AWM represents a significant advancement in the development of intelligent agents capable of learning and applying workflows to improve task-solving efficiency and adaptability in diverse settings.
https://github.com/zorazrw/agent-workflow-memory
Open Interpreter is an open-source natural language interface that enables users to run code locally on their computers through a chat-based system, facilitating tasks like multimedia editing, browser control, and data analysis with full user approval and flexibility beyond hosted solutions.
AIChat is an all-in-one command-line interface tool that integrates multiple large language models and provides features like Shell Assistant, Chat-REPL, RAG, AI Tools, and Agents for enhanced AI-driven command-line productivity.
UFO is a UI-Focused multi-agent framework that automates and navigates Windows OS applications using AI agents powered by GPT-4V to fulfill complex user requests across multiple applications.
FIRERPA/lamda is a robust and versatile Android RPA framework designed for next-generation mobile automation, supporting a wide range of Android versions and environments with extensive APIs and Python SDK for seamless AI integration and device control.
Eko is a production-ready JavaScript framework that enables developers to build reliable, multi-step agentic workflows using natural language commands across computer and browser environments.
Shrimp Task Manager is an intelligent task management system for AI Agents that converts natural language into structured development tasks with dependency tracking, iterative refinement, and research capabilities to enhance programming workflows.
AutoPR is a tool that runs AI-powered workflows over codebases to automate tasks such as documentation, issue tracking, and pull request management.
A curated list of workflow automation software, tools, articles, and resources aimed at improving productivity and efficiency by automating repetitive tasks for individuals and teams.