# Awesome Mcp Open

> MCP server from rizzdev/awesome-mcp-open.

- **Type:** MCP server
- **Install:** `agentstack add mcp-rizzdev-awesome-mcp-open`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [rizzdev](https://agentstack.voostack.com/s/rizzdev)
- **Installs:** 0
- **Category:** [Integrations](https://agentstack.voostack.com/c/integrations)
- **Latest version:** 0.1.0
- **License:** CC0-1.0
- **Upstream author:** [rizzdev](https://github.com/rizzdev)
- **Source:** https://github.com/rizzdev/awesome-mcp-open

## Install

```sh
agentstack add mcp-rizzdev-awesome-mcp-open
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# Awesome MCP Open [](https://awesome.re)

> Model Context Protocol servers you can actually run yourself.

I kept running into the same thing: I'd find a promising MCP server, wire it up, and only then realize it was just a shell around some company's paid API. Cancel the subscription and it's a brick. Most of the big MCP lists are full of these.

So I started keeping my own list, and this is it. Only servers you can actually run yourself. Open source, self-hosted, your data stays on your machine. If a tool only works by calling home to someone else's cloud, it's not here. That's the whole rule, and everything below has been checked against it by hand.

## Contents

- [🔍 Search & Web](#search-web)
- [🌐 Browser Automation](#browser-automation)
- [🗄️ Databases](#databases)
- [📊 Data & Analytics](#data-analytics)
- [🛠️ Developer Tools](#developer-tools)

- [🚢 DevOps & Infra](#devops-infra)
- [☁️ Cloud & Deploy](#cloud-deploy)
- [💳 Payments & Finance](#payments-finance)
- [🎨 Design & Media](#design-media)
- [🧠 Memory](#memory)

- [📋 Productivity & Tasks](#productivity-tasks)
- [💬 Communication](#communication)
- [📣 Social](#social)
- [🗺️ Maps & Location](#maps-location)
- [🔒 Security](#security)

---

### What earns a place

A server has to clear three bars. Miss one and it's out, however good it is otherwise.

- **Self-hostable.** You run it, and nothing in the request path has to touch a vendor's servers.
- **Open source.** A real FOSS license, so you can read the code before you give it access to anything.
- **No lock-in.** It keeps working without a paid backend, and your data isn't stuck on someone else's box.

Plenty gets left out on purpose. Cloud-only servers you can't run. Projects under source-available licenses that aren't really open (n8n's is the usual example). Clients that do nothing until you pay for an API. Mostly good software, just not what I'm collecting here.

---

### Red flags

Some of these work fine but come with a catch you'll want to know about first. Two of them show up in red, and you can confirm either one on the repo page in a few seconds:

| flag | meaning |
|---|---|
| $\textcolor{red}{\textsf{fiddly setup}}$ | wants Docker, a database, or a few services running before it does anything |
| $\textcolor{red}{\textsf{heavy}}$ | throws dozens of tools at the model and eats into your context |

### New & emerging

🌿 means it's newer and still under 500 stars. Often worth a try, just less proven than the ones that have been around a while.

---

## 🔍 Search & Web

- **[Brave Search](https://github.com/brave/brave-search-mcp-server)** · Query Brave's Search, Summarizer, and Local APIs from an MCP server
- **[Crawl4AI](https://github.com/unclecode/crawl4ai)** · Self-hosted LLM-friendly web crawler and scraper with a built-in Docker MCP bridge exposing crawl, scrape, and markdown-extraction tools · $\textcolor{red}{\textsf{fiddly setup}}$
- **[DuckDuckGo](https://github.com/nickclyde/duckduckgo-mcp-server)** · Search DuckDuckGo and fetch/parse page content into clean text with no API key needed
- **[fastCRW](https://github.com/us/crw)** 🌿 · Rust-based self-hostable crawl, scrape, and search server, a keyless Firecrawl/Tavily alternative for agents
- **[Fetch](https://github.com/modelcontextprotocol/servers/tree/main/src/fetch)** · Official reference MCP server that fetches a URL and converts its content to markdown for LLM consumption
- **[Hister](https://github.com/asciimoo/hister)** · Indexes your browsing history and local files into a private, full-text search engine reachable over MCP
- **[Meilisearch](https://github.com/meilisearch/meilisearch-mcp)** · Manage indexes, documents and settings and run searches on a self-hosted Meilisearch instance
- **[Open WebSearch](https://github.com/Aas-ee/open-webSearch)** · Multi-engine web search MCP server scraping Bing, DuckDuckGo, Brave, Startpage and more with no API keys required
- **[SearXNG](https://github.com/ihor-sokoliuk/mcp-searxng)** · Query a self-hosted or public SearXNG metasearch instance for private, ad-free web search results
- **[Wigolo](https://github.com/KnockOutEZ/wigolo)** · Local-first web search, fetch, crawl, and research tools for agents with zero API keys or cloud

## 🌐 Browser Automation

- **[Browser Use](https://github.com/browser-use/browser-use)** · Lets AI agents drive a locally running browser to complete multi-step web tasks, exposed as an MCP server
- **[BrowserTools](https://github.com/AgentDeskAI/browser-tools-mcp)** · Streams live browser console logs, network requests, and screenshots from a Chrome extension into an MCP-compatible IDE · $\textcolor{red}{\textsf{fiddly setup}}$
- **[BrowserWing](https://github.com/browserwing/browserwing)** · Turns recorded browser actions into reusable MCP commands or skills for fast, reliable site automation
- **[Chrome DevTools](https://github.com/ChromeDevTools/chrome-devtools-mcp)** · Controls a local Chrome instance for performance tracing, network inspection, and DOM debugging via the Chrome DevTools Protocol
- **[Computer-Use Agent (CUA)](https://github.com/trycua/cua/tree/main/libs/python/mcp-server)** · Runs a full computer-use agent driving sandboxed macOS/Windows/Linux VMs for cross-OS desktop control · $\textcolor{red}{\textsf{fiddly setup}}$ · $\textcolor{red}{\textsf{heavy}}$
- **[Desktop Commander](https://github.com/wonderwhy-er/DesktopCommanderMCP)** · Gives an agent terminal control, filesystem search, and diff-based file editing on your local desktop
- **[DrissionPage MCP Server](https://github.com/jumodada/Drissionpage-MCP-Server)** 🌿 · Exposes 53 typed DrissionPage browser tools for structured, vision-ready navigation, clicking, and form automation · $\textcolor{red}{\textsf{heavy}}$
- **[OpenDia](https://github.com/aeonfun/opendia)** · Connects AI models to your real Chrome, Firefox, or Arc browser using your existing logins and sessions
- **[Playwright](https://github.com/microsoft/playwright-mcp)** · Drives a real Chromium/Firefox/WebKit browser via Playwright's accessibility tree for navigation, form filling, and scraping
- **[Safari MCP](https://github.com/achiya-automation/safari-mcp)** 🌿 · Automates native Safari on macOS via AppleScript with 80+ tools, keeping logins without a headless browser · $\textcolor{red}{\textsf{heavy}}$
- **[Skyvern](https://github.com/Skyvern-AI/skyvern)** · Automate browser workflows with vision-and-LLM-driven agents instead of brittle selectors
- **[Stealth Browser MCP](https://github.com/vibheksoni/stealth-browser-mcp)** · Drives undetectable real browser instances to bypass Cloudflare and anti-bot systems for AI agents · $\textcolor{red}{\textsf{heavy}}$

## 🗄️ Databases

- **[DBHub](https://github.com/bytebase/dbhub)** · Zero-dependency SQL gateway MCP server for Postgres, MySQL, MariaDB, SQL Server and SQLite with read-only mode and query guardrails
- **[Elasticsearch MCP Server](https://github.com/elastic/mcp-server-elasticsearch)** · Official Elastic server for natural-language search, index management, and cluster analysis against a self-hosted Elasticsearch
- **[FreePeak DB](https://github.com/FreePeak/db-mcp-server)** · Go-based multi-database MCP server supporting concurrent MySQL, PostgreSQL, SQLite, Oracle and TimescaleDB connections with schema and performance tools · $\textcolor{red}{\textsf{heavy}}$
- **[Milvus](https://github.com/zilliztech/mcp-server-milvus)** · Official MCP server exposing collection search, insert, and query tools for a self-hosted Milvus vector database · $\textcolor{red}{\textsf{fiddly setup}}$
- **[MindsDB](https://github.com/mindsdb/mindsdb)** · Query and federate data across many databases and sources through a self-hosted MCP-enabled server · $\textcolor{red}{\textsf{fiddly setup}}$
- **[MongoDB MCP Server](https://github.com/mongodb-js/mongodb-mcp-server)** · Official server for querying, managing, and aggregating self-hosted MongoDB or Atlas clusters through natural language
- **[MotherDuck / DuckDB](https://github.com/motherduckdb/mcp-server-motherduck)** · Runs SQL analytics over local, in-memory, or S3-hosted DuckDB files with no MotherDuck account required
- **[MySQL](https://github.com/designcomputer/mysql_mcp_server)** · Enables secure, configurable read/write SQL access to a self-hosted MySQL database
- **[Neo4j](https://github.com/neo4j-contrib/mcp-neo4j)** · Run Cypher queries and manage graph data and schema on a self-hosted Neo4j database
- **[OpenSearch MCP Server](https://github.com/opensearch-project/opensearch-mcp-server-py)** 🌿 · Official OpenSearch Project server exposing search, index, and cluster-management tools for self-hosted OpenSearch clusters
- **[Postgres MCP Pro](https://github.com/crystaldba/postgres-mcp)** · Runs against your own Postgres to give schema introspection, EXPLAIN-based query tuning, index advice, and configurable read/write SQL access
- **[Qdrant](https://github.com/qdrant/mcp-server-qdrant)** · Official MCP server for storing and semantically searching data in a self-hosted Qdrant vector database · $\textcolor{red}{\textsf{fiddly setup}}$
- **[Redis](https://github.com/redis/mcp-redis)** · Natural-language interface to manage, query, and inspect data in a Redis instance · $\textcolor{red}{\textsf{fiddly setup}}$
- **[Supabase](https://github.com/supabase/mcp)** · Manage Supabase projects, tables, and queries from an MCP server
- **[Toolbox for Databases](https://github.com/googleapis/mcp-toolbox)** · Connect an agent to Postgres, MySQL, Spanner, and other databases through a shared open toolbox
- **[Weaviate](https://github.com/weaviate/weaviate)** · Vector database with a native MCP endpoint you enable and self-host via a single environment variable · $\textcolor{red}{\textsf{fiddly setup}}$

## 📊 Data & Analytics

- **[Apache Airflow MCP Server](https://github.com/yangkyeongmo/mcp-server-apache-airflow)** 🌿 · Exposes Apache Airflow DAGs, runs, and task instances as MCP tools for orchestrating self-hosted data pipelines
- **[Apache Pinot MCP Server](https://github.com/startreedata/mcp-pinot)** 🌿 · StarTree-maintained server for querying and inspecting self-hosted Apache Pinot real-time OLAP analytics clusters
- **[Apache Superset](https://github.com/apache/superset/tree/master/superset/mcp_service)** · Built-in service letting AI agents query and manage dashboards, charts, and datasets on a self-hosted Superset instance · $\textcolor{red}{\textsf{fiddly setup}}$
- **[ClickHouse](https://github.com/ClickHouse/mcp-clickhouse)** · Official MCP server for running read/write SQL queries and browsing schema on a self-hosted ClickHouse cluster · $\textcolor{red}{\textsf{fiddly setup}}$
- **[dbt](https://github.com/dbt-labs/dbt-mcp)** · Official server running dbt Core CLI commands, codegen, and local lineage/manifest introspection for a dbt project
- **[Grafana](https://github.com/grafana/mcp-grafana)** · Query and manage self-hosted Grafana dashboards, datasources, alerts, and incidents via its API · $\textcolor{red}{\textsf{heavy}}$
- **[Grafana Loki](https://github.com/incu6us/loki-mcp-server)** · Discover labels and run LogQL queries against a self-hosted Grafana Loki instance
- **[Metabase MCP](https://github.com/jerichosequitin/metabase-mcp)** 🌿 · Bridges AI assistants to a self-hosted Metabase, querying dashboards and running SQL questions
- **[OpenTelemetry (Traceloop)](https://github.com/traceloop/opentelemetry-mcp-server)** · Query and analyze OpenTelemetry traces from self-hosted Jaeger or Tempo backends for debugging and LLM observability
- **[Prometheus](https://github.com/pab1it0/prometheus-mcp-server)** · Query and analyze metrics from a self-hosted Prometheus server using PromQL
- **[Trino MCP Server](https://github.com/tuannvm/mcp-trino)** 🌿 · High-performance Go server letting AI agents query and explore self-hosted Trino data warehouses

## 🛠️ Developer Tools

- **[ast-grep](https://github.com/ast-grep/ast-grep-mcp)** · Exposes ast-grep structural code search and codemod rewriting as MCP tools
- **[Commands](https://github.com/g0t4/mcp-server-commands)** · Run arbitrary shell commands or direct executables on the host machine via a runProcess tool
- **[DevDocs](https://github.com/cyberagiinc/DevDocs)** · Crawl and index any tech documentation site into a private, searchable MCP knowledge base · $\textcolor{red}{\textsf{fiddly setup}}$
- **[Filesystem](https://github.com/modelcontextprotocol/servers/tree/main/src/filesystem)** · Read, write, edit, move, search, and manage files and directories on the local filesystem with configurable root access control
- **[Firefox DevTools MCP](https://github.com/mozilla/firefox-devtools-mcp)** 🌿 · Official Mozilla server exposing Firefox remote debugging to AI assistants for live page inspection
- **[Ghidra MCP](https://github.com/bethington/ghidra-mcp)** · 200+ tools for AI-driven reverse engineering through Ghidra, as a GUI plugin plus a headless server · $\textcolor{red}{\textsf{heavy}}$
- **[Git](https://github.com/modelcontextprotocol/servers/tree/main/src/git)** · Official reference MCP server exposing local git operations like status, diff, commit, and log
- **[GitHub](https://github.com/github/github-mcp-server)** · Self-hosted server exposing GitHub repository, issue, PR, and code-search operations to MCP clients · $\textcolor{red}{\textsf{heavy}}$
- **[Language Server (LSP)](https://github.com/isaacphi/mcp-language-server)** · Wraps any language server to expose LSP definitions, references, rename, and diagnostics as MCP tools
- **[MarkItDown](https://github.com/microsoft/markitdown)** · Convert PDFs, Office docs, images, and more to markdown for LLM consumption
- **[OpenAPI](https://github.com/ivo-toby/mcp-openapi-server)** · Turn any local or remote OpenAPI specification into callable MCP tools, prompts, and resources
- **[Pandoc](https://github.com/vivekVells/mcp-pandoc)** · Convert documents between Markdown, HTML, PDF, DOCX, and other formats using local Pandoc · $\textcolor{red}{\textsf{fiddly setup}}$
- **[Sequential Thinking](https://github.com/modelcontextprotocol/servers/tree/main/src/sequentialthinking)** · Structures dynamic, revisable step-by-step reasoning for complex problem solving
- **[Serena](https://github.com/oraios/serena)** · Provides LSP-based semantic code retrieval and editing tools (find symbol, references, rename) across languages · $\textcolor{red}{\textsf{heavy}}$
- **[Shell Server](https://github.com/tumf/mcp-shell-server)** · Execute a whitelisted set of shell commands with stdin support through a secure argv-based MCP interface
- **[x64dbg MCP](https://github.com/SetsunaYukiOvO/x64dbg-mcp)** 🌿 · Plugin bridging the x64dbg/x32dbg debugger to AI agents over JSON-RPC for remote binary debugging

## 🚢 DevOps & Infra

- **[Azure DevOps MCP](https://github.com/microsoft/azure-devops-mcp)** · Official Microsoft server bringing Azure DevOps boards, repos, and pipelines into AI coding agents
- **[Containarium](https://github.com/FootprintAI/Containarium)** 🌿 · Self-hostable agent runtime giving each AI agent an SSH-isolated sandbox on Kubernetes or LXC
- **[Docker Hub MCP](https://github.com/docker/hub-mcp)** 🌿 · Official Docker server for image discovery, search, and repository management via LLMs
- **[Helm](https://github.com/zekker6/mcp-helm)** · Runs Helm CLI operations (install, upgrade, list, repo management) against a Kubernetes cluster · $\textcolor{red}{\textsf{fiddly setup}}$
- **[Kubernetes](https://github.com/containers/kubernetes-mcp-server)** · Native Go MCP server for Kubernetes and OpenShift cluster management with no external CLI dependency
- **[Kubernetes (Flux159)](https://github.com/Flux159/mcp-server-kubernetes)** · Manages Kubernetes clusters via kubectl/helm, covering pods, deployments, services, and manifests · $\textcolor{red}{\textsf{heavy}}$
- **[RootCause](https://github.com/yindia/rootcause)** 🌿 · Local-first server turning natural-l

…

## Source & license

This open-source MCP server is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [rizzdev](https://github.com/rizzdev)
- **Source:** [rizzdev/awesome-mcp-open](https://github.com/rizzdev/awesome-mcp-open)
- **License:** CC0-1.0

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** yes
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/mcp-rizzdev-awesome-mcp-open
- Seller: https://agentstack.voostack.com/s/rizzdev
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.