MCP AI apps

Integration platform to connect AI agents and products to 800+ external APIs.

Nango is an integration platform for building and running product and agent integrations across 800+ APIs. You write integration logic as TypeScript—or let AI generate it—and deploy to Nango's runtime, which handles auth, execution, scaling, and observability. AI agents consume those integrations through an MCP server, tool calls, webhooks, and data syncs. It runs as managed cloud or self-hosted.

Open-source integration gateway that connects AI agents to tools via MCP.

Metorial is an open-source integration and access layer for AI agents, exposing 1,200+ tools through the Model Context Protocol with consistent auth, permissions, and session observability. It can host custom MCP servers on serverless infrastructure and provides Python and TypeScript SDKs, a hosted cloud, and self-hosting. Every MCP session is recorded for review in its dashboard.

Zero-dependency, token-efficient database MCP server.

An open-source Model Context Protocol server that lets AI coding assistants connect to and query databases through one unified gateway. DBHub deliberately exposes only two tools — execute_sql (with transaction support and safety controls) and search_objects (schema/table/column exploration) — to keep the agent's context window lean. It supports Postgres, MySQL, SQL Server, MariaDB, and SQLite, works with clients like Claude Desktop, Claude Code, Cursor, VS Code, and Copilot CLI, and includes a built-in web interface for running queries and viewing request traces.

AI-ready iPaaS to build agents, govern MCP, and automate across apps.

Enterprise integration and automation platform that layers AI agents and governed Model Context Protocol (MCP) onto a mature iPaaS. Teams build conversational or autonomous agents in the Merlin agent builder, expose 700+ app connectors as governed MCP tools through an Agent Gateway, and automate processes with visual low-code or headless code-first builds — all under role-based access, audit, and compliance controls.

Inspector to test, debug, and evaluate MCP servers and ChatGPT apps.

MCPJam is a local-first developer tool, a 'Postman for MCP', for testing, debugging, inspecting, and evaluating Model Context Protocol servers. You can manually run tools, resources, resource templates, prompts, and elicitation flows with full JSON-RPC observability, debug every message and OAuth exchange, and run evaluations across multiple LLMs to catch regressions. It runs as a hosted web app, a Mac/Windows desktop app, or via npx in your terminal.

Build, curate, and host Model Context Protocol servers from your APIs or TypeScript.

Gram is a platform for creating, curating, and hosting Model Context Protocol (MCP) servers. Instead of exposing raw API endpoints that confuse agents, it lets you compose higher-order, business-level tools from an OpenAPI document or TypeScript functions, group them into toolsets, and deploy each as a hosted, authenticated MCP server. Every toolset is instantly usable from Claude, ChatGPT, Cursor, or any MCP-compatible client, with an optional OAuth 2.1 proxy and org-wide access controls.

Unified API for B2B integrations — plus MCP tools for AI agents.

Merge offers a single Unified API that adds hundreds of product integrations (HR, ATS, CRM, accounting, ticketing, file storage) behind one normalized schema. Its Agent Handler product exposes those same integrations as production-ready MCP connectors, giving AI agents thousands of pre-built, authenticated tools. Engineering teams use it to ship integrations and agent tool-calling without hand-building each connector.

MCP-native eval and observability server for AI agents.

Iris scores AI agent output, catches safety failures, and enforces cost budgets — exposed as an MCP server rather than an SDK. Any MCP-compatible agent discovers its tools (trace logging, output evaluation, rule management, LLM-as-judge) and uses them automatically, with no code changes, so every output flowing through the protocol gets evaluated. It detects PII leaks, prompt injection, hallucinations, and budget anomalies. The core is MIT-licensed and free to self-host; a managed cloud adds dashboards and alerting.

Open-source toolkit to connect any LLM to any MCP server.

mcp-use is an open-source framework for building MCP clients, servers, and agents in TypeScript and Python. It lets you connect any LLM to any Model Context Protocol server in a few lines of code — configuring servers, choosing a model, and spinning up a tool-using agent — without depending on a closed application client. On top of the official SDK it adds opinionated developer ergonomics: hot reload, dev tools, auth providers, and UI components.

Aggregate many MCP servers into one secured gateway.

MetaMCP is an open-source gateway that aggregates multiple Model Context Protocol (MCP) servers into unified, namespaced endpoints. Running as a single Docker service, it lets you group servers, apply middleware to filter or rewrite the tools they expose, and serve everything over SSE or streamable HTTP to any MCP client. It adds API-key and OAuth authentication, multi-tenancy, and a built-in MCP inspector for debugging.

Open-source MCP gateway, registry, and proxy.

ContextForge is an open-source gateway that sits in front of Model Context Protocol servers, federating them — alongside A2A agents and REST/gRPC APIs — behind a single endpoint with centralized governance. It speaks HTTP, JSON-RPC, WebSocket, SSE, stdio, and streamable HTTP, and ships an admin UI, OpenTelemetry-based observability, Redis-backed caching, and multi-cluster federation. Built and maintained by IBM, it targets production deployments that need to manage many MCP tools at once.

Run and manage containerized MCP servers from Docker Desktop.

Docker's Model Context Protocol stack: a Catalog of 200+ verified, containerized MCP servers (from publishers like Stripe, GitHub, Grafana, MongoDB, and Notion), a Toolkit built into Docker Desktop for one-click install and secret/OAuth management, and an open-source Gateway that proxies many servers behind a single secure endpoint. Each server runs in its own isolated container, so agents in Claude Desktop, Cursor, or VS Code get controlled tool access without hand-editing JSON config.

MCP integrations and live environments for AI agents.

Klavis AI is an open-source layer of Model Context Protocol (MCP) servers and clients that lets AI agents reliably use external tools like GitHub, Gmail, Google Sheets, Jira, and Slack. It handles OAuth and multi-tenant auth, ships hosted servers plus self-hostable images, and has expanded into managed sandbox environments for training and evaluating long-horizon agents. Teams can call the hosted endpoints or run the open-source servers on their own infra.

Persistent memory layer for AI coding agents.

ByteRover gives coding agents long-term, structured memory that persists across sessions and tools. It organizes project knowledge into hierarchical context trees, stores them locally by default, and serves relevant context back to agents through an MCP server and CLI. It plugs into Cursor, Claude Code, Windsurf, and other assistants, and an optional cloud tier lets teams sync and share context.

Build and deploy AI agents as APIs, with MCP tools built in.

Toolhouse is a backend-as-a-service for AI agents: define an agent in plain language or code and deploy it as an API in one command. Agents connect automatically to the Toolhouse MCP server, which provides RAG, memory, code execution, browser use, and 1,000+ tool integrations over the Model Context Protocol.

Hand-reviewed directory and news hub for the Model Context Protocol.

PulseMCP is a discovery hub for the Model Context Protocol (MCP) ecosystem — a daily-updated, hand-reviewed directory of thousands of MCP servers and clients, validated by live handshake and scored on reliability. Alongside the directory it publishes tutorials, a use-case showcase, an API for programmatic access, and 'The Agentic Loop' newsletter. It is a free resource for finding and tracking MCP integrations.

Community directory for discovering Model Context Protocol servers.

mcp.so is a community-driven directory that collects and organizes third-party MCP (Model Context Protocol) servers so developers can browse, share, and integrate external tools for AI apps. It ranks servers by real call volume with a "Call Ranking" leaderboard and offers Featured, Latest, Official, and Hosted filters. The project is open-source and still in preview.

Enterprise platform to discover, govern, and protect AI agents and applications.

An enterprise AI security platform that inventories every model, agent, MCP server, and data dependency across your stack, scans for misconfigurations and supply-chain risks, red-teams AI applications, and enforces runtime guardrails against prompt injection and rogue agent behavior. Founded in Tel Aviv in 2023, it counts UiPath, Best Buy, and Nielsen among its customers. Enterprise pricing via demo.

Open-source MCP server giving AI clients terminal and file control.

Desktop Commander is an open-source Model Context Protocol server that lets MCP clients like Claude Desktop, Cursor, and VS Code run terminal commands, search the file system, and make surgical diff-based edits. It manages long-running processes, reads and writes files (including Excel and PDF), and turns an AI assistant into a hands-on coding and automation agent on your own machine. Free and MIT-licensed, with one of the larger community followings among MCP servers.

Connect APIs, AI, and databases with code-level workflows.

A workflow automation platform where steps are pre-built triggers/actions or arbitrary Node.js and Python code, glued across 3,000+ managed app integrations with hosted auth. Pipedream Connect embeds those integrations into your own product or AI agent, and the catalog is exposed MCP-natively so agents can call any connector as a tool.

Open-source, AI-first workflow automation — a self-hostable Zapier alternative.

A no-code automation platform with a visual drag-and-drop builder, 400+ app integrations, and native AI steps. It runs as a managed cloud or fully self-hosted on your own infrastructure, and its pieces double as MCP tools that AI agents can call.

MCP server registry, inspector, and gateway.

A discovery and hosting hub for the Model Context Protocol ecosystem: browse a large indexed catalog of MCP servers, test them in an in-browser Inspector, and route them through a managed Gateway that handles credentials, logging, and analytics. Browsing and installing open-source servers locally is free; hosting servers on Glama's infrastructure and the Gateway's managed features are paid. Also ships an AI playground chat client over the connected tools.

Up-to-date, version-specific library docs piped into any AI coding agent via MCP.

An MCP server from Upstash that pulls current, version-specific documentation and code examples for a library and feeds them straight into an LLM's context. It exposes two tools — resolve-library-id and a docs query — so agents in Cursor, Claude Code, Windsurf, and other MCP clients generate code against the real, current API.

Microsoft's official MCP server for browser automation, driven by Playwright.

An open-source Model Context Protocol server from Microsoft that lets LLMs and agents browse and act on the web through Playwright. It exposes 40+ tools for navigation, forms, clicking, network mocking, tracing, and more, operating on structured accessibility snapshots rather than pixels so each interactive element gets a deterministic ref. Works with Claude, VS Code, Cursor, Windsurf, and any MCP client.

The MCP runtime for secure, authenticated agent tool-calling.

A platform that lets AI agents take real, authenticated actions across business systems like Gmail, GitHub, Slack, and Notion. Arcade handles the hard part of agent tool-use — secure, per-user authorization — so a tool call runs with the actual end user's permissions rather than a shared API key. It ships agent-optimized prebuilt tools plus an open framework for building custom ones, exposed over MCP. A free tier is available for getting started.

Registry and hosting platform for Model Context Protocol servers.

A central hub for the MCP ecosystem: browse, install, and configure thousands of community-built MCP servers from one place, or deploy your own to Smithery's infrastructure and access it remotely. The CLI installs and manages servers without hand-editing JSON config, and the hosted tier adds managed OAuth and persistent connections so agents get authenticated tool access. Lists over 6,000 servers across developer tools, data connectors, and productivity integrations.

The fast, Pythonic way to build MCP servers and clients.

An open-source Python framework for building Model Context Protocol servers and clients. You declare a tool as a Python function and FastMCP generates the schema, validation, and protocol plumbing, and handles transport, auth, and the connection lifecycle for you. FastMCP 1.0 was folded into the official MCP Python SDK; v2 is the actively developed standalone superset. The framework is free and Apache-2.0; Prefect's Horizon platform adds managed hosting and enterprise governance.

Authenticated toolkits that give AI agents access to SaaS apps, via MCP or API.

Tool layer that gives AI agents secure, managed access to 1000+ toolkits and 20,000+ tools across SaaS apps, exposed via MCP or direct APIs. It handles authentication, tool search, and context management, and its MCP Gateway hands each team a managed endpoint to paste into Claude, Cursor, or ChatGPT. The core is MIT-licensed; a free tier covers 20K tool calls per month.

Google's open-source Gemini AI agent, in your terminal.

An open-source command-line agent that brings Gemini into the terminal for coding, debugging, content generation, and research. Ships built-in Google Search grounding, Model Context Protocol support, and project context via a GEMINI.md file, plus a non-interactive mode for scripting and CI. Free to use by signing in with a personal Google account, with paid options through AI Studio, Vertex AI, and Gemini Code Assist.

On-machine AI agent for coding, workflows, and automation.

An extensible AI agent from Block that runs natively on your machine as a desktop app or CLI, executing code and automating engineering tasks via MCP extensions. It supports 15+ LLM providers (Anthropic, OpenAI, Google, Ollama, and more) using your own keys. Free and open source under Apache 2.0.

Reads + writes GitHub from any MCP-aware agent.

Official GitHub MCP server. Wired into Claude Code and most agentic surfaces to read PRs, post comments, create branches, and run secret-scanning.

Postgres + auth + edge functions from inside the agent.

MCP server that exposes Supabase project ops — list tables, apply migrations, deploy edge functions, fetch advisors. Removes a context-switch cliff between code and database.