Connect a client

Every client below reaches the same two things: the MCP server at https://mcp.mzizi.dev/mcp, and the agent skills. Each section gives the vendor’s own config format, checked against that vendor’s documentation on 20 October 2026 and linked, so you can check it again. Formats move; where a vendor’s page and this one disagree, the vendor’s page wins.

What you connect

Endpoint	Use it for
`https://mcp.mzizi.dev/mcp`	Every tool, with no account. Streamable HTTP. It never asks you to sign in
`https://mcp.mzizi.dev/mcp/signin`	The same server behind sign-in. It answers `401` until you sign in, then serves every tool
`npx -y @nyuchi/mzizi-mcp` (stdio, on your machine)	Every tool except `mzizi_fundi`, and the bundled data, offline except the `docs_*` tools. No credential

Which tools need sign-in. Only mzizi_fundi and mzizi_report_issue, which act as your Mzizi console user. Connected to /mcp, they are listed, and calling one answers with how to sign in. A client starts its OAuth flow only when a server answers 401, and /mcp never does, so to use those two tools, add /mcp/signin (as a second server, or instead of /mcp). The server supports dynamic client registration, so a client that asks how to register should register automatically. It does not publish a client ID metadata document. How the skills arrive. One of three ways, depending on the client:

Natively, from the mzizi plugin. Claude Code, and Codex and ChatGPT through the same plugin, install the seven skills as skills and connect the MCP server in one step.
Natively, from npm. npm install -D @nyuchi/mzizi-skills, then npx skills experimental_sync, links the skills into .agents/skills/, which Codex, Cursor, VS Code, Gemini CLI and OpenCode all read.
Over MCP. Any client connected to the server can call mzizi_get_skills: with no arguments it lists the skills and their categories, name returns one in full, and category (dev, design, language or build) returns every skill in it in full. Nothing to install. This is the route for Claude.ai, ChatGPT without the plugin, the APIs and Windsurf.

At a glance

Client	Connect with	Skills arrive	For the Fundi tools
Claude.ai	A custom connector	Over MCP	A second connector at `/mcp/signin`
Claude Desktop	A custom connector, or `claude_desktop_config.json` (stdio)	Over MCP	The connector route; stdio has no `mzizi_fundi`
Claude Code	The `mzizi` plugin, or `claude mcp add`	Natively with the plugin; over MCP otherwise	`claude mcp add` for `/mcp/signin`, then `/mcp`
The Claude API	The MCP connector (`mcp_servers`)	Over MCP	`/mcp/signin` with an `authorization_token`
ChatGPT	A custom MCP server under Plugins	Over MCP	A second server at `/mcp/signin`
OpenAI Codex	The `mzizi` plugin, or `config.toml`	Natively with the plugin or `.agents/skills/`; over MCP otherwise	A `/mcp/signin` table, then `codex mcp login`
The OpenAI API and Agents SDK	A `type: "mcp"` tool	Over MCP	`/mcp/signin` with `authorization`
Z.ai’s GLM Coding Plan	Claude Code (or another tool it supports)	As that tool gets them	As that tool does
OpenCode	`opencode.json`	Natively from `.agents/skills/`, or over MCP	A second `remote` entry, then `opencode mcp auth`
Cursor	`mcp.json`	Natively from `.agents/skills/`, or over MCP	A second entry at `/mcp/signin`
VS Code and GitHub Copilot	`MCP: Add Server`, `.mcp.json` or `mcp-config.json`	Natively from `.agents/skills/`, or over MCP	A second HTTP server at `/mcp/signin`
Windsurf (Cascade)	`mcp_config.json`	Over MCP	A second `serverUrl` entry at `/mcp/signin`
Gemini CLI	`gemini mcp add` or `settings.json`	Natively from `.agents/skills/`, or over MCP	`gemini mcp add` for `/mcp/signin`

Claude

Claude.ai

Add a custom connector: in Customize > Connectors, choose + Add, then Add custom connector, name it Mzizi, and enter https://mcp.mzizi.dev/mcp. Under Authentication, choose No sign in. On Team and Enterprise plans an owner adds it once in Organization settings > Connectors (Add, then Custom, then Web), and each member connects to it. For the Fundi tools, add a second connector at https://mcp.mzizi.dev/mcp/signin, choose a sign-in option under Authentication, and Register automatically under OAuth client. Enable it in a conversation from +, then Connectors. Skills arrive over MCP: ask Claude to call mzizi_get_skills. Free plans are limited to one custom connector. Claude connects to the server from Anthropic’s cloud, not from your device; mcp.mzizi.dev is public, so that works. Checked against Get started with custom connectors using remote MCP.

Claude Desktop

The custom connector above works in Claude Desktop too; it is configured on your Claude account, not on the machine. To run the server on your machine instead, over stdio, open Settings from the Claude menu, then Developer, then Edit Config, and add it to claude_desktop_config.json (~/Library/Application Support/Claude/ on macOS, %APPDATA%\Claude\ on Windows):

{
  "mcpServers": {
    "mzizi": {
      "command": "npx",
      "args": ["-y", "@nyuchi/mzizi-mcp"]
    }
  }
}

Restart Claude Desktop. The stdio server needs Node.js, holds no credential, and does not offer mzizi_fundi. Skills arrive over MCP. Checked against the custom connectors article above and Connect to local MCP servers.

Claude Code

The plugin carries the seven skills and connects the server:

/plugin marketplace add mzizi-dev/mzizi-registry
/plugin install mzizi@mzizi

From a shell, the same is claude plugin marketplace add mzizi-dev/mzizi-registry and claude plugin install mzizi@mzizi. It lives in the plugin/ directory of mzizi-dev/mzizi-registry; see the skills bundle for which skills version it carries. Or the server alone, with skills over MCP:

claude mcp add --transport http --scope user mzizi https://mcp.mzizi.dev/mcp
# for the Fundi tools; run /mcp in a session to sign in
claude mcp add --transport http --scope user mzizi-signin https://mcp.mzizi.dev/mcp/signin

Leave out --scope user to add it to the current project only. On 20 October 2026, claude mcp list showed mzizi as connected and mzizi-signin as needing authentication, and the plugin installed from a local copy of the registry listed seven skills and the mzizi server, connected. Checked against the Claude Code MCP, plugin manifest and marketplace references.

The Claude API

The Messages API’s MCP connector reaches a remote server for you. It needs the mcp-client-2025-11-20 beta header, a server in mcp_servers and an mcp_toolset that names it. Use the model you would otherwise call:

curl https://api.anthropic.com/v1/messages \
  -H "content-type: application/json" \
  -H "x-api-key: $ANTHROPIC_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "anthropic-beta: mcp-client-2025-11-20" \
  -d '{
    "model": "YOUR_MODEL",
    "max_tokens": 2048,
    "messages": [{"role": "user", "content": "Load the Mzizi dev skills."}],
    "mcp_servers": [
      {"type": "url", "url": "https://mcp.mzizi.dev/mcp", "name": "mzizi"}
    ],
    "tools": [{"type": "mcp_toolset", "mcp_server_name": "mzizi"}]
  }'

No authorization_token is needed for /mcp. Only tool calls are supported, so skills arrive through mzizi_get_skills. The Fundi tools need a token from the sign-in flow in authorization_token, against /mcp/signin; the connector does not run that flow for you. Checked against MCP connector. Not run here: it needs an API key.

OpenAI

ChatGPT

Go to ChatGPT Plugins, select +, then Add custom MCP server. Name it Mzizi, enter https://mcp.mzizi.dev/mcp under Connection, set no authentication, review the warning, and select Create as a plugin. Install it, start a new conversation, and type @ to pick it. Account and workspace policies decide whether you can add a custom server. Skills arrive over MCP. The mzizi plugin also packages the skills for ChatGPT, in the portable layout below. The ChatGPT desktop app reads a repository’s .agents/plugins/marketplace.json, but installing it from GitHub has not been tested here. Checked against Connect and test your plugin.

OpenAI Codex

The plugin. The registry’s plugin/ carries a portable Agent Plugins manifest (plugin/plugin.json and plugin/mcp.json) beside the Claude one, and a repo marketplace at .agents/plugins/marketplace.json:

codex plugin marketplace add mzizi-dev/mzizi-registry

Then install mzizi from the Plugins directory: in the ChatGPT desktop app, or with /plugins in the Codex CLI where your version has it. Start a new session after installing. The plugin’s mcp.json connects /mcp over streamable-http.

The Codex and ChatGPT manifests arrive in mzizi-dev/mzizi-registry with its pull request #520. Until that reaches the registry’s main branch, codex plugin marketplace add finds no plugin there: use the server alone, below.

Or the server alone, in ~/.codex/config.toml (or a trusted project’s .codex/config.toml), which the Codex CLI, the IDE extension and the ChatGPT desktop app share:

[mcp_servers.mzizi]
url = "https://mcp.mzizi.dev/mcp"

# for the Fundi tools; then run: codex mcp login mzizi-signin
[mcp_servers.mzizi-signin]
url = "https://mcp.mzizi.dev/mcp/signin"

codex mcp add mzizi --url https://mcp.mzizi.dev/mcp writes the first table for you. Skills. Codex reads skills from .agents/skills/ in the repository and from ~/.agents/skills/, which is where npx skills experimental_sync links them; otherwise they arrive over MCP. Codex also reads AGENTS.md before any work, so a repository can tell it to load the skills:

## Mzizi

Before any work, call `mzizi_get_skills` with `category: "dev"` on the Mzizi MCP server
(`https://mcp.mzizi.dev/mcp`), and load `mzizi-roots`, `mzizi-design` or `mzizi-language`
before touching components, styling or `.mz` source.

Checked against Package your plugin, Model Context Protocol, Build skills and Custom instructions with AGENTS.md. The manifests validate against the Agent Plugins schemas at agent-plugins.org/schemas/1.0.0. Not run here: the Codex CLI was not installed.

The OpenAI API and Agents SDK

In the Responses API, the server is a type: "mcp" tool. Use the model you would otherwise call:

{
  "model": "YOUR_MODEL",
  "tools": [
    {
      "type": "mcp",
      "server_label": "mzizi",
      "server_description": "The Mzizi language, components, tokens, doctrine, skills and docs.",
      "server_url": "https://mcp.mzizi.dev/mcp",
      "require_approval": "never"
    }
  ],
  "input": "Load the Mzizi dev skills, then summarise them."
}

Every tool on /mcp is marked read-only except the two Fundi tools, which refuse without sign-in, and docs_submit_feedback, which sends feedback about a docs page. Keep "require_approval": "never" only if you accept that, or narrow the list with allowed_tools. For the Fundi tools, point server_url at /mcp/signin and pass an access token in authorization. In the Agents SDK for Python, the same object goes in a hosted tool, and the Responses API makes the connection:

from agents import Agent, HostedMCPTool

agent = Agent(
    name="Mzizi builder",
    instructions="Load the Mzizi dev skills with mzizi_get_skills before any work.",
    tools=[
        HostedMCPTool(
            tool_config={
                "type": "mcp",
                "server_label": "mzizi",
                "server_url": "https://mcp.mzizi.dev/mcp",
                "require_approval": "never",
            }
        )
    ],
)

To connect from your own process instead, use MCPServerStreamableHttp with params={"url": "https://mcp.mzizi.dev/mcp"} and pass it in mcp_servers. Skills arrive over MCP. Checked against MCP servers and the Agents SDK’s MCP page. Not run here: it needs an API key.

Z.ai’s GLM Coding Plan

Z.ai’s coding plan runs inside coding tools rather than as its own client. Its documented Claude Code setup points Claude Code at Z.ai with ANTHROPIC_BASE_URL set to https://api.z.ai/api/anthropic and your Z.ai key in ANTHROPIC_AUTH_TOKEN, in ~/.claude/settings.json, or through Z.ai’s helper, npx @z_ai/coding-helper. Plugins and MCP servers are Claude Code’s own, so connect Mzizi exactly as in Claude Code: the mzizi plugin, or claude mcp add. Z.ai’s own MCP servers use the same commands (claude mcp add -s user …). Z.ai documents other tools too. In OpenCode, Cursor or another client, follow that client’s section here. Checked against Z.ai’s Claude Code and Vision MCP server pages. Not run here: it needs a Z.ai key.

Editors and other CLIs

OpenCode

In opencode.json (the project’s, or your global config):

{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "mzizi": {
      "type": "remote",
      "url": "https://mcp.mzizi.dev/mcp",
      "enabled": true
    }
  }
}

For the Fundi tools, add a second remote entry at /mcp/signin. OpenCode starts OAuth itself when a server answers 401, or run opencode mcp auth <name>. Skills: OpenCode reads .agents/skills/ and ~/.agents/skills/, so the npm route works; otherwise over MCP. Checked against MCP servers and Agent Skills.

Cursor

In .cursor/mcp.json in the project, or ~/.cursor/mcp.json for every project:

{
  "mcpServers": {
    "mzizi": {
      "url": "https://mcp.mzizi.dev/mcp"
    }
  }
}

Add a mzizi-signin entry at /mcp/signin for the Fundi tools; Cursor runs the OAuth flow. Skills: Cursor reads .agents/skills/ (and .cursor/skills/, and Claude’s and Codex’s skill folders), so the npm route works; otherwise over MCP. Checked against Cursor’s MCP and Skills pages.

VS Code and GitHub Copilot

Run MCP: Add Server from the Command Palette, choose HTTP, enter https://mcp.mzizi.dev/mcp and the ID mzizi. Or from a shell:

code --add-mcp '{"name":"mzizi","type":"http","url":"https://mcp.mzizi.dev/mcp"}'

To share it with a team, commit a workspace-root .mcp.json. Its portable format, which GitHub Copilot CLI also reads from ~/.copilot/mcp-config.json, uses mcpServers:

{
  "mcpServers": {
    "mzizi": {
      "type": "http",
      "url": "https://mcp.mzizi.dev/mcp"
    }
  }
}

VS Code’s own .vscode/mcp.json uses servers instead of mcpServers, with the same entry. Skills: VS Code reads .agents/skills/, .github/skills/ and .claude/skills/, so the npm route works; otherwise over MCP. Checked against Add and manage MCP servers, the MCP configuration reference and Agent skills.

Windsurf

Windsurf’s documentation now sits under Devin Desktop. For the Cascade agent, open the … menu in the Cascade panel, then the MCP config file, and add the server to mcp_config.json (~/.config/devin/mcp_config.json on macOS and Linux, %APPDATA%\devin\mcp_config.json on Windows). A remote server takes serverUrl:

{
  "mcpServers": {
    "mzizi": {
      "serverUrl": "https://mcp.mzizi.dev/mcp"
    }
  }
}

Skills arrive over MCP. The page says the Devin Local agent, now the default for new tabs, keeps MCP servers in the Devin CLI’s config instead; that format is not covered here. Checked against Cascade MCP configuration.

Gemini CLI

gemini mcp add --transport http --scope user mzizi https://mcp.mzizi.dev/mcp

That writes ~/.gemini/settings.json; without --scope user it writes the project’s .gemini/settings.json. By hand, a Streamable HTTP server takes httpUrl:

{
  "mcpServers": {
    "mzizi": {
      "httpUrl": "https://mcp.mzizi.dev/mcp"
    }
  }
}

Skills: Gemini CLI reads .agents/skills/ and ~/.agents/skills/, so the npm route works; otherwise over MCP. Checked against MCP servers with Gemini CLI and Agent skills.

Any other MCP client

Most clients take one of the shapes above: a Streamable HTTP URL, https://mcp.mzizi.dev/mcp, or the stdio command npx -y @nyuchi/mzizi-mcp. The server is listed in the MCP Registry as io.github.mzizi-dev/mzizi-mcp, so a client that installs from the registry finds both.

What is not verified here

The Codex plugin in a live Codex CLI or ChatGPT desktop app. The manifests follow OpenAI’s packaging guide and validate against the Agent Plugins schemas, but no Codex install ran.
Per-server sign-in in the portable mcp.json. OpenAI’s guide puts OAuth settings under a server’s extensions["com.openai"].auth, but the Agent Plugins MCP schema allows no extensions key on a server. The plugin’s mcp.json connects the anonymous /mcp and declares no auth; connect /mcp/signin yourself for the Fundi tools.
The Devin Local agent’s MCP config, in Windsurf’s successor. Only the Cascade agent’s mcp_config.json is documented above.
The Claude API, the OpenAI API and Z.ai snippets come from the vendors’ pages; none was run, because each needs an account key.

​What you connect

​At a glance

​Claude

​Claude.ai

​Claude Desktop

​Claude Code

​The Claude API

​OpenAI

​ChatGPT

​OpenAI Codex

​The OpenAI API and Agents SDK

​Z.ai’s GLM Coding Plan

​Editors and other CLIs

​OpenCode

​Cursor

​VS Code and GitHub Copilot

​Windsurf

​Gemini CLI

​Any other MCP client

​What is not verified here

What you connect

At a glance

Claude

Claude.ai

Claude Desktop

Claude Code

The Claude API

OpenAI

ChatGPT

OpenAI Codex

The OpenAI API and Agents SDK

Z.ai’s GLM Coding Plan

Editors and other CLIs

OpenCode

Cursor

VS Code and GitHub Copilot

Windsurf

Gemini CLI

Any other MCP client

What is not verified here