Skip to content
agentzone.one

status: launching soon

Find the right AI agent or MCP server, and set it up correctly.

AgentZone is a curated directory of AI agents and Model Context Protocol servers, organized by category, with setup guides and compatibility notes for Claude, ChatGPT, Cursor, and more.

Get early access What is MCP? →
~/agentzone/directory preview 
$ tree --categories
.
├── coding/
├── research/
├── data/
├── browser-automation/
├── productivity/
└── infra/

# each entry will include:
  what-it-does   tools   access
  setup          compat  security-notes

/categories

Browse by what you need done

Listings are being reviewed now. These are the categories the directory will open with.

  • 01 / coding indexing

    Coding

    Agents and servers that read repositories, run tests, search code, and work with issue trackers and pull requests.

    • repo search
    • test runners
    • issue trackers
    • CI logs
  • 02 / research indexing

    Research

    Web search, page fetching, documentation lookup, and tools that help an agent gather and cite sources.

    • web search
    • docs lookup
    • paper search
    • citations
  • 03 / data indexing

    Data

    Database and analytics connectors for querying, inspecting schemas, and transforming data, ideally read-only by default.

    • SQL databases
    • spreadsheets
    • warehouses
    • schemas
  • 04 / browser indexing

    Browser automation

    Servers that let an agent navigate pages, click, fill forms, and take screenshots for sites without an API.

    • navigation
    • screenshots
    • form filling
    • scraping
  • 05 / productivity indexing

    Productivity

    Connections to docs, calendars, chat, notes, and project tools for drafting, summarizing, and follow-up.

    • docs & wikis
    • calendars
    • team chat
    • tasks
  • 06 / infra indexing

    Infra & DevOps

    Cloud, container, and observability tools, listed with extra care around permissions and write access.

    • containers
    • cloud APIs
    • logs
    • monitoring

compatibility notes planned for:

  • Claude
  • ChatGPT
  • Cursor
  • VS Code
  • Windsurf
  • Zed
  • CLI agents

/how-it-works

From “what should I use?” to a working setup

  1. 01

    Browse by category

    Start from what you want to get done: coding, research, data, browser automation, or productivity. Each category groups agents and MCP servers that solve related problems.

  2. 02

    Check fit and compatibility

    Each entry will say what it does, which tools it exposes, what it can access, and which AI clients it is known to work with.

  3. 03

    Follow the setup guide

    Copy a config snippet for your client, set the required environment variables, and test it, with security notes alongside.

/features

What every listing is designed to answer

[curated] Curated, not scraped
Entries are reviewed for maintenance and clarity before they are listed. Fewer, better options beat an endless list.
[access] Plain-language access notes
What each server can read, write, or send, so you can apply least privilege before you install.
[compat] Client compatibility
Notes on which clients an entry works with, and where configuration differs between them.
[setup] Copy-ready setup
Launch commands and config shapes for local (stdio) and remote (HTTP) servers, with common pitfalls.
[type] Agents and servers, side by side
Complete agents and the MCP servers that extend them, in one place, clearly labeled.
[free] Readable without an account
The directory is planned as a public, static site. No sign-up wall to read a listing.

/use-cases

Who it is for

you are you want to start with
A developer Give your coding assistant access to issues, docs, and databases coding, data
A small team Automate status reports, triage, and follow-ups without building infrastructure productivity
A researcher or analyst Gather sources, compare findings, and query data with citations research, data
An ops or growth person Handle repetitive work in web tools that have no API browser-automation
A tool builder See what already exists before writing your own agent or server all categories

/launch

The directory is launching soon

We are reviewing agents and MCP servers category by category and writing setup guides for each. There is no sign-up list yet. When the first categories are ready, they will appear right here on agentzone.one.

in the meantime

/faq

Questions

What is AgentZone?

AgentZone is a curated directory of AI agents and Model Context Protocol (MCP) servers, organized by category, with setup guides and client compatibility notes. It is not live yet; this site and blog are the first part.

What is an MCP server?

An MCP server is a program that exposes tools, data, or prompts to AI applications using the open Model Context Protocol. Any MCP-capable client can connect to it. Our article on what MCP is explains the details.

Which AI clients will listings cover?

We plan to note compatibility with widely used MCP-capable clients such as Claude apps, ChatGPT, Cursor, VS Code, and command-line coding agents. Support for MCP features varies by client, so notes will be specific where it matters.

When does the directory launch?

Soon. We are reviewing entries and writing setup guides now. We are not publishing a date until the first categories are ready.

Will it cost anything?

Reading the directory is planned to be free. We will not ask for an account to view listings.

Is it safe to connect agents to my tools?

It can be, with care: use scoped credentials, require approval for writes, and treat anything an agent reads from the web or email as untrusted. Our security checklist covers the essentials.

/blog

Latest guides

All articles →