Frameworks, platforms, and libraries for building agentic systems, deploying AI applications, and managing ML workflows.
A library for building stateful, multi-actor applications with LLMs. Provides fine-grained control over agent workflows with support for cycles, branching, and persistence.
Framework for orchestrating role-playing autonomous AI agents. Enables collaborative intelligence where agents work together with defined roles, goals, and delegation capabilities.
A framework for building multi-agent conversational AI systems. Supports diverse conversation patterns with customizable and conversable agents for complex LLM workflows.
OpenAI's lightweight SDK for building agentic applications. Provides primitives for agent loops, tool use, handoffs between agents, and guardrails for production deployments.
TypeScript toolkit for building AI-powered applications. Provides React hooks, streaming utilities, and adapters for all major LLM providers with a unified API.
The fastest way to build and share data apps. Turns Python scripts into interactive web applications with minimal code — ideal for ML demos and internal tools.
Build machine learning demos and web applications with a few lines of Python. Supports image, text, audio inputs and integrates directly with Hugging Face Spaces.
Cloud platform for running generative AI models, large-scale batch jobs, and interactive applications. Provides serverless GPU infrastructure with a Python-native developer experience.
Run and deploy machine learning models via a cloud API. Host open-source models or deploy your own with automatic scaling and a simple prediction API.
ML experiment tracking, dataset versioning, and model management platform. Track hyperparameters, visualize metrics, and collaborate on ML projects with your team.
Open-source platform for the complete ML lifecycle. Provides experiment tracking, reproducible runs, model packaging, and a central model registry for deployment.
Data Version Control — Git for data and ML models. Track datasets, pipelines, and experiments with Git-like commands. Supports storage on S3, GCS, Azure, and more.
The platform for sharing and discovering ML models, datasets, and spaces. Host models with version control, automatic model cards, and inference APIs.
Anthropic's reference MCP server implementations including Filesystem, Git, Memory, Fetch, and Sequential Thinking. The canonical starting point for adopting the Model Context Protocol.
Pulls current, version-specific documentation for thousands of libraries directly into your LLM context at query time. Fetches actual docs instead of relying on outdated training data.
Microsoft's MCP server for end-to-end browser automation, testing, and web scraping. Uses structured accessibility trees for faster, lighter interactions than screenshot-based approaches.
Web-based marketplace for discovering, installing, and managing MCP servers. Offers one-click configuration snippets for various editors and AI clients.
Community-curated directory of MCP servers across categories — databases, developer tools, productivity, cloud, and more. The most popular MCP resource with 83K+ GitHub stars.
The official community-driven registry of MCP servers. A searchable app store for discovering and listing MCP servers with verified metadata.