{"type":"mcp_client","name":"AWorld","description":"AWorld facilitates building, evaluating, and running General Multi-Agent Assistance, bridging the gap between theoretical MAS capabilities and real-world applications. It allows users to create agentic prototypes and extend them for practical needs.","category":"AI","language":"Python","stars":1204,"forks":3,"owner":"inclusionAI","github_url":"https://github.com/inclusionAI/AWorld","homepage":null,"setup":"## Setup\nWith Python>=3.11:\n```bash\npython setup.py install\n```\n\n## Environment Configuration\n```bash\n# Choose your preferred AI model(s) and set the corresponding API key(s)\n# OpenAI (Required for GPT-3.5, GPT-4)\nexport OPENAI_API_KEY=sk-abcd1234wxyz5678...\n# Anthropic Claude (Required for Claude 2, Claude 3)\nexport CLAUDE_API_KEY=sk-ant-api03xyz...\n```","tools":"## Available Tools\n\n1.  Browsers (Chrome, Firefox) (Controls web browsers for navigation, form filling, and interaction with web pages)\n2.  Android device simulation (Manages Android device simulation for mobile app testing and automation)\n3.  Shell (Executes shell commands for file operations and system interactions)\n4.  Code (Runs code snippets in various languages for data processing and automation)\n5.  Search (Performs web searches and returns structured results for information gathering and summary)\n6.  Document (Handles file operations including reading, writing, and managing directories)\n7.  Agent initialization (AI-Powered Agents)\n8.  Delegation between multiple agents (AI-Powered Agents)\n","faq":null,"created_at":"2025-03-14T08:30:52+00:00","updated_at":"2025-03-28T12:32:16+00:00","source_url":"https://model-context-protocol.com/clients/build-evaluate-run-multi-agent-assistance","related_articles":[]}