awesome-production-genai

Agent2Agent (A2A) protocol addresses a critical challenge in the AI landscape: enabling gen AI agents, built on diverse frameworks by different companies running on separate servers, to communicate and collaborate effectively - as agents, not just as tools.

The Agent Client Protocol (ACP) standardizes communication between code editors (interactive programs for viewing and editing source code) and coding agents (programs that use generative AI to autonomously modify code).

Control Claude Code, AmazonQ, Opencode, Goose, Aider, Gemini, GitHub Copilot, Sourcegraph Amp, Codex, Auggie, and Cursor CLI with an HTTP API.

The agents.json Specification is an open specification that formally describes contracts for API and agent interactions, built on top of the OpenAPI standard.

AgentNetworkProtocol (ANP) is an open-source communication standard designed to enable seamless connectivity and collaboration between intelligent agents, positioning itself as the foundational protocol for agent-to-agent interactions in the emerging AI ecosystem.

AP2 (Agent Payments Protocol) is a protocol for building secure and interoperable AI-driven payments in agentic commerce workflows.

A2UI is a framework for building agent-to-UI interactions.

Chat UI is an open-source web application framework that provides the frontend interface and backend infrastructure for building conversational AI chatbots, serving as the codebase behind their HuggingChat platform.

ComfyUI is a node-based interface and inference engine for generative AI, specifically designed to work with Stable Diffusion and other AI models. It allows users to create complex workflows for image, video, and other content generation through a visual, graph-like interface.

ADK is Google's Agent Development Kit for Python, a framework for building production-ready AI agents.

AgentKit help agent developers build multi-agent networks with deterministic routing and rich tooling via MCP.

Agent Lightning is a framework for building production-ready AI agents with Lightning AI.

Agents allows users to build AI-driven server programs that can see, hear, and speak in realtime.

AgentScope is a multi-agent platform designed to empower developers to build multi-agent applications with large-scale models.

Agentset is an open-source production-ready RAG platform with built-in agentic reasoning, hybrid search, and multimodal support.

Agentic Radar is a security scanner for LLM agentic workflows for potential vulnerabilities.

Agentic Security is a vulnerability scanner for agentic workflows, protecting AI systems from jailbreaks, fuzzing, and multimodal attacks.

AI-Infra-Guard is an infrastructure security tool for AI systems.

DeepTeam is a simple-to-use, open-source LLM red teaming framework, for penetration testing and safe guarding large-language model systems.

Open-source AI agent security scanner. Detects prompt injection, infinite loops, token bombing, SQL injection via LLM, and missing human oversight across 20+ frameworks. CLI + MCP server with compliance mapping to EU AI Act, NIST AI RMF, and OWASP LLM Top 10.

mcp-scan is an MCP security scanning tool for local and remote MCP Servers.

AgentLab is an open-source framework for developing, testing, and benchmarking web agents on diverse tasks, designed for scalability and reproducibility.

AgentOps helps developers build, evaluate, and monitor AI agents from prototype to production.

IntellAgent is an advanced multi-agent framework that transforms the evaluation and optimization of conversational agents.

Judgeval is an open-source framework for agent behavior monitoring. Judgeval offers a toolkit to track and judge agent behavior in online and offline setups, enabling you to convert interaction data from production/test environments into improved agents.

Manifest is open-source observability for AI agents. Track costs, tokens, messages, and performance — entirely on your machine.

ChainForge is an open-source visual programming environment for battle-testing prompts to LLMs. Compare across models, prompts, and prompt parameters using built-in visualizations.

DSPy is the framework for programming—not prompting—language models. It allows you to iterate fast on building modular AI systems and offers algorithms for optimizing prompts and weights.

ell is a language model programming library that treats prompts as programs. Features automatic versioning, serialization, and rich tooling for prompt engineering with Ell Studio for visualization.

Latitude is the open-source prompt engineering platform to build, evaluate, and refine prompts with AI. Features prompt management, playground testing, AI gateway, and evaluations.

PromptIDE by xAI is an integrated development environment for prompt engineering and interpretability research, providing transparent access to Grok-1 with rich analytics and Python SDK support.

PromptLayer is a platform that allows you to track, manage, and share your GPT prompt engineering by acting as middleware to log all OpenAI API requests.

Graphiti is a framework for building and querying temporally-aware knowledge graphs, specifically tailored for AI agents operating in dynamic environments.

LangMem provides ways to extract meaningful details from chats, store them, and use them to improve future interactions.

Mem0 ("mem-zero") enhances AI assistants and agents with an intelligent memory layer, enabling personalized AI interactions. It remembers user preferences, adapts to individual needs, and continuously learns over time—ideal for customer support chatbots, AI assistants, and autonomous systems.

MemOS is an operating system for Large Language Models (LLMs) that enhances them with long-term memory capabilities. It allows LLMs to store, retrieve, and manage information, enabling more context-aware, consistent, and personalized interactions.

SimpleMem is an efficient lifelong memory system for LLM agents that enables seamless memory retention and management across conversations.

supermemory intelligently extracts information from your conversations and apps and pieces together connections between memories to deliver seamless user experience.