# Agent Field Guide

> Agent Field Guide is an independent directory of AI agent tools: agent frameworks, MCP servers, coding agents, agent apps, automation platforms and agent infrastructure. Each entry links to the official site.

## Agent Frameworks

- [LangGraph](https://agentfieldguide.net/tools/langgraph/): Low-level orchestration framework from LangChain for building stateful, multi-step agents as graphs, with persistence and human-in-the-loop support. Official: https://www.langchain.com/langgraph
- [LangChain](https://agentfieldguide.net/tools/langchain/): Popular framework for building LLM applications and agents with composable chains, tools, retrievers and integrations. Official: https://www.langchain.com
- [CrewAI](https://agentfieldguide.net/tools/crewai/): Python framework for orchestrating role-playing, collaborative multi-agent "crews" with tasks and processes. Official: https://www.crewai.com
- [AutoGen](https://agentfieldguide.net/tools/autogen/): Microsoft's open-source framework for building multi-agent conversational applications. Official: https://github.com/microsoft/autogen
- [OpenAI Agents SDK](https://agentfieldguide.net/tools/openai-agents-sdk/): Lightweight OpenAI SDK for multi-agent workflows with handoffs, guardrails and tracing. Official: https://github.com/openai/openai-agents-python
- [LlamaIndex](https://agentfieldguide.net/tools/llamaindex/): Data framework for connecting LLMs to your data, with agent and workflow abstractions for RAG-heavy agents. Official: https://www.llamaindex.ai
- [Semantic Kernel](https://agentfieldguide.net/tools/semantic-kernel/): Microsoft SDK for integrating LLMs into apps with plugins, planners and agent abstractions. Official: https://github.com/microsoft/semantic-kernel
- [Pydantic AI](https://agentfieldguide.net/tools/pydantic-ai/): Agent framework from the Pydantic team focused on type-safe, structured outputs and dependency injection. Official: https://ai.pydantic.dev
- [smolagents](https://agentfieldguide.net/tools/smolagents/): Hugging Face's minimal library for agents that think and act in code. Official: https://github.com/huggingface/smolagents
- [Mastra](https://agentfieldguide.net/tools/mastra/): TypeScript agent framework with workflows, memory, RAG, evals and integrations. Official: https://mastra.ai
- [Vercel AI SDK](https://agentfieldguide.net/tools/vercel-ai-sdk/): TypeScript toolkit for building AI apps and agents with streaming, tool calling and multi-provider support. Official: https://ai-sdk.dev
- [Agno](https://agentfieldguide.net/tools/agno/): Python framework (formerly Phidata) for building multi-modal agents with memory, knowledge and tools. Official: https://www.agno.com
- [Haystack](https://agentfieldguide.net/tools/haystack/): deepset's open-source framework for production RAG pipelines and agents. Official: https://haystack.deepset.ai
- [Google Agent Development Kit](https://agentfieldguide.net/tools/google-adk/): Google's open-source toolkit for building, evaluating and deploying agents, optimized for Gemini. Official: https://google.github.io/adk-docs/
- [DSPy](https://agentfieldguide.net/tools/dspy/): Stanford framework for programming—rather than prompting—language models, with optimizers for prompts and weights. Official: https://dspy.ai
- [CAMEL](https://agentfieldguide.net/tools/camel-ai/): Open-source multi-agent framework and research community studying agent societies. Official: https://www.camel-ai.org
- [Letta](https://agentfieldguide.net/tools/letta/): Framework (from the MemGPT team) for stateful agents with long-term memory. Official: https://www.letta.com
- [Claude Agent SDK](https://agentfieldguide.net/tools/claude-agent-sdk/): Anthropic's SDK for building agents on the same harness that powers Claude Code. Official: https://docs.claude.com/en/api/agent-sdk/overview

## MCP Servers

- [Filesystem MCP Server](https://agentfieldguide.net/tools/mcp-filesystem/): Reference MCP server for secure file operations with configurable access controls. Official: https://github.com/modelcontextprotocol/servers/tree/main/src/filesystem
- [Fetch MCP Server](https://agentfieldguide.net/tools/mcp-fetch/): Reference MCP server that fetches web pages and converts them to markdown for LLMs. Official: https://github.com/modelcontextprotocol/servers/tree/main/src/fetch
- [Git MCP Server](https://agentfieldguide.net/tools/mcp-git/): Reference MCP server to read, search and manipulate Git repositories. Official: https://github.com/modelcontextprotocol/servers/tree/main/src/git
- [Memory MCP Server](https://agentfieldguide.net/tools/mcp-memory/): Reference MCP server providing a knowledge-graph-based persistent memory. Official: https://github.com/modelcontextprotocol/servers/tree/main/src/memory
- [GitHub MCP Server](https://agentfieldguide.net/tools/github-mcp/): GitHub's official MCP server for repos, issues, PRs and Actions. Official: https://github.com/github/github-mcp-server
- [Playwright MCP](https://agentfieldguide.net/tools/playwright-mcp/): Microsoft's MCP server for browser automation via Playwright accessibility snapshots. Official: https://github.com/microsoft/playwright-mcp
- [Supabase MCP Server](https://agentfieldguide.net/tools/supabase-mcp/): Connects AI assistants to Supabase projects for tables, queries and config. Official: https://github.com/supabase-community/supabase-mcp
- [Stripe MCP / Agent Toolkit](https://agentfieldguide.net/tools/stripe-mcp/): Stripe's agent toolkit and MCP server for interacting with the Stripe API. Official: https://github.com/stripe/agent-toolkit
- [Notion MCP Server](https://agentfieldguide.net/tools/notion-mcp/): Official Notion MCP server for reading and writing Notion pages and databases. Official: https://github.com/makenotion/notion-mcp-server
- [Cloudflare MCP Servers](https://agentfieldguide.net/tools/cloudflare-mcp/): Cloudflare's MCP servers for Workers, observability, DNS and more. Official: https://github.com/cloudflare/mcp-server-cloudflare
- [Sentry MCP Server](https://agentfieldguide.net/tools/sentry-mcp/): Sentry's MCP server for querying issues, errors and performance data. Official: https://github.com/getsentry/sentry-mcp
- [Context7](https://agentfieldguide.net/tools/context7/): Upstash MCP server that injects up-to-date, version-specific library docs into prompts. Official: https://github.com/upstash/context7
- [Firecrawl MCP Server](https://agentfieldguide.net/tools/firecrawl-mcp/): MCP server for web scraping, crawling and search via Firecrawl. Official: https://github.com/mendableai/firecrawl-mcp-server
- [Brave Search MCP Server](https://agentfieldguide.net/tools/brave-search-mcp/): Brave's MCP server for web, news and local search via the Brave Search API. Official: https://github.com/brave/brave-search-mcp-server
- [Exa MCP Server](https://agentfieldguide.net/tools/exa-mcp/): MCP server for Exa's AI-native web search API. Official: https://github.com/exa-labs/exa-mcp-server
- [Postgres MCP Pro](https://agentfieldguide.net/tools/mcp-postgres/): Postgres MCP server with query execution, index tuning and health checks. Official: https://github.com/crystaldba/postgres-mcp
- [Linear MCP Server](https://agentfieldguide.net/tools/linear-mcp/): Linear's official remote MCP server for issues, projects and comments. Official: https://linear.app/docs/mcp
- [Browserbase MCP Server](https://agentfieldguide.net/tools/browserbase-mcp/): Cloud browser automation for agents via Browserbase and Stagehand. Official: https://github.com/browserbase/mcp-server-browserbase
- [Figma Dev Mode MCP](https://agentfieldguide.net/tools/figma-mcp/): Figma's MCP server exposing design context to coding agents. Official: https://help.figma.com/hc/en-us/articles/32132100833559
- [Zapier MCP](https://agentfieldguide.net/tools/zapier-mcp/): Connects agents to thousands of apps and actions through Zapier. Official: https://zapier.com/mcp
- [Sequential Thinking MCP Server](https://agentfieldguide.net/tools/mcp-sequential-thinking/): Reference MCP server that helps a model work through problems with structured, revisable step-by-step reasoning. Official: https://github.com/modelcontextprotocol/servers/tree/main/src/sequentialthinking
- [Time MCP Server](https://agentfieldguide.net/tools/mcp-time/): Reference MCP server for current time and timezone conversions. Official: https://github.com/modelcontextprotocol/servers/tree/main/src/time
- [Everything MCP Server](https://agentfieldguide.net/tools/mcp-everything/): Reference test server exercising every MCP feature (prompts, resources, tools), useful for client developers. Official: https://github.com/modelcontextprotocol/servers/tree/main/src/everything
- [Chrome DevTools MCP](https://agentfieldguide.net/tools/chrome-devtools-mcp/): Official Chrome DevTools MCP server letting coding agents control and debug a live Chrome browser. Official: https://github.com/ChromeDevTools/chrome-devtools-mcp
- [AWS MCP Servers](https://agentfieldguide.net/tools/aws-mcp/): AWS Labs' suite of MCP servers for AWS documentation, infrastructure and services. Official: https://github.com/awslabs/mcp
- [Azure MCP Server](https://agentfieldguide.net/tools/azure-mcp/): Microsoft's MCP servers, including Azure, for managing cloud resources from agents. Official: https://github.com/microsoft/mcp
- [MongoDB MCP Server](https://agentfieldguide.net/tools/mongodb-mcp/): Official MongoDB MCP server for querying databases and managing Atlas. Official: https://github.com/mongodb-js/mongodb-mcp-server
- [Neon MCP Server](https://agentfieldguide.net/tools/neon-mcp/): Manage Neon serverless Postgres projects and run SQL via natural language. Official: https://github.com/neondatabase/mcp-server-neon
- [Atlassian Remote MCP Server](https://agentfieldguide.net/tools/atlassian-mcp/): Atlassian's remote MCP server for Jira and Confluence. Official: https://www.atlassian.com/platform/remote-mcp-server
- [Slack MCP Server](https://agentfieldguide.net/tools/slack-mcp/): Popular community MCP server for reading and searching Slack workspaces. Official: https://github.com/korotovsky/slack-mcp-server
- [Grafana MCP Server](https://agentfieldguide.net/tools/grafana-mcp/): Grafana's MCP server for dashboards, datasources and incident data. Official: https://github.com/grafana/mcp-grafana
- [HubSpot MCP Server](https://agentfieldguide.net/tools/hubspot-mcp/): HubSpot's MCP server for CRM data access by AI agents. Official: https://developers.hubspot.com/mcp

## Coding Agents

- [Cursor](https://agentfieldguide.net/tools/cursor/): AI code editor with agent mode, background agents and codebase-aware chat. Official: https://cursor.com
- [Claude Code](https://agentfieldguide.net/tools/claude-code/): Anthropic's agentic coding tool for the terminal and IDE. Official: https://www.anthropic.com/claude-code
- [OpenAI Codex](https://agentfieldguide.net/tools/openai-codex/): OpenAI's software engineering agent available in ChatGPT, CLI and IDE. Official: https://openai.com/codex/
- [GitHub Copilot](https://agentfieldguide.net/tools/github-copilot/): GitHub's AI pair programmer with agent mode and coding agent. Official: https://github.com/features/copilot
- [Windsurf](https://agentfieldguide.net/tools/windsurf/): Agentic IDE with the Cascade agent. Official: https://windsurf.com
- [Cline](https://agentfieldguide.net/tools/cline/): Open-source autonomous coding agent for VS Code. Official: https://cline.bot
- [Aider](https://agentfieldguide.net/tools/aider/): AI pair programming in your terminal, git-aware. Official: https://aider.chat
- [Devin](https://agentfieldguide.net/tools/devin/): Cognition's autonomous AI software engineer. Official: https://devin.ai
- [Replit Agent](https://agentfieldguide.net/tools/replit-agent/): Builds and deploys full apps from natural language inside Replit. Official: https://replit.com/ai
- [Bolt.new](https://agentfieldguide.net/tools/bolt/): StackBlitz's in-browser agent for prompting, running and deploying full-stack apps. Official: https://bolt.new
- [Lovable](https://agentfieldguide.net/tools/lovable/): Builds full-stack web apps from chat prompts. Official: https://lovable.dev
- [v0](https://agentfieldguide.net/tools/v0/): Vercel's generative UI and app builder. Official: https://v0.app
- [OpenHands](https://agentfieldguide.net/tools/openhands/): Open-source platform for software development agents (formerly OpenDevin). Official: https://github.com/All-Hands-AI/OpenHands
- [Gemini CLI](https://agentfieldguide.net/tools/gemini-cli/): Google's open-source AI agent bringing Gemini to the terminal. Official: https://github.com/google-gemini/gemini-cli
- [Roo Code](https://agentfieldguide.net/tools/roo-code/): Open-source AI coding agent for VS Code with custom modes (a Cline fork). Official: https://roocode.com
- [Kilo Code](https://agentfieldguide.net/tools/kilo-code/): Open-source AI coding agent for VS Code and JetBrains. Official: https://kilocode.ai
- [Zed](https://agentfieldguide.net/tools/zed/): High-performance code editor with an agent panel and Agent Client Protocol support. Official: https://zed.dev
- [Continue](https://agentfieldguide.net/tools/continue/): Open-source AI code assistant and agents for VS Code and JetBrains. Official: https://www.continue.dev
- [Amp](https://agentfieldguide.net/tools/amp/): Sourcegraph's agentic coding tool for the terminal and editors. Official: https://ampcode.com
- [Jules](https://agentfieldguide.net/tools/jules/): Google's asynchronous coding agent that works on GitHub repos in the cloud. Official: https://jules.google
- [opencode](https://agentfieldguide.net/tools/opencode/): Open-source AI coding agent built for the terminal. Official: https://opencode.ai
- [Augment Code](https://agentfieldguide.net/tools/augment-code/): AI coding agent with a context engine for large codebases. Official: https://www.augmentcode.com
- [Junie](https://agentfieldguide.net/tools/junie/): JetBrains' AI coding agent inside JetBrains IDEs. Official: https://www.jetbrains.com/junie/

## Agent Apps

- [Manus](https://agentfieldguide.net/tools/manus/): General-purpose autonomous agent that executes multi-step tasks. Official: https://manus.im
- [ChatGPT Agent](https://agentfieldguide.net/tools/chatgpt-agent/): ChatGPT's agent mode that browses, uses tools and completes tasks on a virtual computer. Official: https://openai.com/index/introducing-chatgpt-agent/
- [Perplexity Comet](https://agentfieldguide.net/tools/perplexity-comet/): Agentic browser from Perplexity. Official: https://www.perplexity.ai/comet
- [Genspark](https://agentfieldguide.net/tools/genspark/): AI agent workspace for research, slides, sheets and calls. Official: https://www.genspark.ai

## No-code & Automation

- [n8n](https://agentfieldguide.net/tools/n8n/): Fair-code workflow automation platform with AI agent nodes. Official: https://n8n.io
- [Zapier Agents](https://agentfieldguide.net/tools/zapier-agents/): Build AI agents that work across Zapier's app ecosystem. Official: https://zapier.com/agents
- [Make](https://agentfieldguide.net/tools/make/): Visual automation platform with AI agents. Official: https://www.make.com
- [Relevance AI](https://agentfieldguide.net/tools/relevance-ai/): No-code platform for building AI agent workforces. Official: https://relevanceai.com
- [Lindy](https://agentfieldguide.net/tools/lindy/): No-code AI assistants/agents for email, scheduling and workflows. Official: https://www.lindy.ai
- [Dify](https://agentfieldguide.net/tools/dify/): Open-source LLM app platform with visual agent and workflow builder. Official: https://dify.ai
- [Flowise](https://agentfieldguide.net/tools/flowise/): Open-source low-code builder for agents and LLM flows. Official: https://flowiseai.com
- [Langflow](https://agentfieldguide.net/tools/langflow/): Visual, open-source builder for agents and RAG apps. Official: https://www.langflow.org

## Agent Infrastructure

- [LangSmith](https://agentfieldguide.net/tools/langsmith/): Observability, evaluation and tracing platform for LLM apps and agents. Official: https://www.langchain.com/langsmith
- [Langfuse](https://agentfieldguide.net/tools/langfuse/): Open-source LLM engineering platform: tracing, evals, prompt management. Official: https://langfuse.com
- [AgentOps](https://agentfieldguide.net/tools/agentops/): Monitoring and session replay for AI agents. Official: https://www.agentops.ai
- [E2B](https://agentfieldguide.net/tools/e2b/): Secure cloud sandboxes for running AI-generated code. Official: https://e2b.dev
- [Browserbase](https://agentfieldguide.net/tools/browserbase/): Headless browser infrastructure for AI agents. Official: https://www.browserbase.com
- [Mem0](https://agentfieldguide.net/tools/mem0/): Memory layer for AI agents and assistants. Official: https://mem0.ai
- [Composio](https://agentfieldguide.net/tools/composio/): Tool integrations and auth layer for AI agents. Official: https://composio.dev
- [Tavily](https://agentfieldguide.net/tools/tavily/): Search API built for AI agents and RAG. Official: https://www.tavily.com

## Guides

- [Best MCP Servers for Cursor](https://agentfieldguide.net/guides/best-mcp-servers-for-cursor/): MCP servers that make Cursor's agent much more useful: current docs, browser control, GitHub, databases and error tracking. Add them to Cursor's MCP settings.
- [Best Open-Source AI Agent Frameworks](https://agentfieldguide.net/guides/best-open-source-agent-frameworks/): Open-source frameworks for building production AI agents, from graph orchestration to multi-agent crews and type-safe Python.
- [Best Open-Source AI Coding Agents](https://agentfieldguide.net/guides/best-open-source-coding-agents/): Open-source coding agents you can run with your own API keys, in the terminal or in VS Code.
- [Best AI Coding Agents](https://agentfieldguide.net/guides/best-ai-coding-agents/): AI coding agents read your codebase, edit multiple files, run commands and iterate on the result. The right one depends less on raw model quality and more on where you want the agent to work: inside an editor, in your terminal, or in a cloud sandbox that hands back a pull request.
- [Cursor vs Claude Code vs Codex](https://agentfieldguide.net/guides/cursor-vs-claude-code-vs-codex/): Cursor, Claude Code and OpenAI Codex are three of the most popular AI coding agents. They overlap more every release, but each still starts from a different place: Cursor is an AI-first editor, Claude Code is Anthropic's terminal agent, and Codex is OpenAI's agent that spans a CLI, IDE extension and cloud tasks.
- [Best MCP Servers](https://agentfieldguide.net/guides/best-mcp-servers/): MCP (Model Context Protocol) servers give AI assistants and agents tools and data: your repos, a real browser, databases, docs and the apps your team uses. These are the servers that add the most value, grouped by use case, with links to the official sources.
- [Cursor Alternatives](https://agentfieldguide.net/guides/cursor-alternatives/): Cursor is a popular AI code editor, but it is not the only way to get agentic coding. Some developers want to keep their current editor, some prefer a terminal agent, and some want open-source tools with their own API keys. These are the strongest alternatives, grouped by why you might switch.
- [Best AI App Builders](https://agentfieldguide.net/guides/best-ai-app-builders/): AI app builders, sometimes called vibe coding tools, turn a plain-language description into a running web app with hosting included. They suit founders, designers and developers who want a prototype or MVP fast. Here is how the main options differ and what they cost.
- [Zapier Alternatives for AI Agents](https://agentfieldguide.net/guides/zapier-alternatives-for-ai-agents/): Zapier is the default for no-code automation, but teams building AI agents often want self-hosting, cheaper high-volume runs, visual LLM flows or agents that act more like teammates. These alternatives cover each of those needs.

## About

- [About](https://agentfieldguide.net/about/)
- [Submit a tool](https://agentfieldguide.net/submit/)
- [Sponsor](https://agentfieldguide.net/sponsor/)