mcc-fleet
A wrapper MCP server for Minecraft Console Client's MCP server. Allowing for multiple bot instances with per-bot action
Install / Use
claude mcp add italoseara -- npx -y github:italoseara/mcc-fleetIf the server publishes to npm under a different name, use that package instead — check the repo README.
MCP Server
Model Context Protocol server
Quality Score
Category
Development & EngineeringSupported Platforms
Skill content
View source on GitHubmcc-fleet
An MCP server that manages a fleet of Minecraft Console Client (MCC) instances and lets an MCP client (e.g. Claude) drive each bot individually.
MCC already ships an embedded MCP server, but each mcc process controls only one
account. This wrapper:
- Spawns/stops multiple
mccprocesses — each with its own nick, working directory, and MCP HTTP port. - Proxies the in-game tools of each MCC instance, addressed by nick.
Targets offline-mode servers (login by nick, no Microsoft auth) and MCC v26.1.
How it works
Claude -- stdio --> mcc-fleet (this wrapper) -- HTTP /mcp --> mcc (Bot1)
-- HTTP /mcp --> mcc (Bot2)
...
Each bot's MCC config is rendered from mcc-template.ini into runtime/<nick>/MinecraftClient.ini.
MCC's embedded MCP endpoint only comes up after the bot joins the world, so spawn_bot
polls it until ready.
Tools exposed to the client
| Tool | Purpose |
|------|---------|
| spawn_bot(nick, host?, port?, timeout?) | Start a bot, wait until ready, return its tools |
| list_bots() | Status, MCP port, pid, uptime of all bots |
| list_bot_tools(nick) | Discover a bot's in-game tools |
| bot_call(nick, tool, arguments) | Invoke one of a bot's in-game tools |
| stop_bot(nick) / stop_all() | Disconnect/terminate bot(s) |
Dependencies
| Dependency | Why | Version |
|---|---|---|
| Python | Runs the wrapper | 3.12+ |
| uv | Installs Python deps and runs the server | any recent |
| Minecraft Console Client (mcc) | The bot client each instance wraps | v26.1 |
mcc is a self-contained native binary (no separate .NET runtime needed) and must be
reachable on PATH as mcc, or pointed to via MCC_BINARY (see below).
Installation
-
Install
uv(skip if already installed):curl -LsSf https://astral.sh/uv/install.sh | sh -
Install Minecraft Console Client v26.1. Download the build for your platform from the releases page, then make it executable and put it on
PATHasmcc:chmod +x MinecraftClient sudo mv MinecraftClient /usr/local/bin/mcc(If you'd rather not move it onto
PATH, leave it where it is and setMCC_BINARYto its full path in the next step instead.) -
Clone this repo and install Python dependencies:
git clone https://github.com/italoseara/mcc-fleet.git cd mcc-fleet uv sync -
Configure the target server (defaults:
localhost:25565):export MCC_SERVER_HOST=play.example.net export MCC_SERVER_PORT=25565 # optional: export MCC_BINARY=mcc # path to the mcc executable, if not on PATH export MCC_BASE_MCP_PORT=33334 # first MCP port to allocate -
Sanity check — the wrapper should start without errors:
uv run mcc-wrapperIt should print a FastMCP startup banner and then sit waiting for MCP messages on stdio;
Ctrl+Cto exit. This confirmsuv syncand themccbinary are both set up correctly.
Register with Claude Code
claude mcp add mcc-fleet \
-e MCC_SERVER_HOST=play.example.net -e MCC_SERVER_PORT=25565 \
-- uv run --directory /path/to/mcc-fleet mcc-wrapper
Then ask Claude to spawn_bot("Bot1"), bot_call("Bot1", ...), etc.
Notes
- The server must be in offline-mode for nick-only login.
- On shutdown the wrapper kills all child
mccprocesses (stop_all).
Related Skills
momen-cursurrules-prompt-file
40.6kCursor rules for building custom frontends with Momen.app as headless BaaS with GraphQL API, actionflows, AI agents, and Stripe integration.
semiotic-react-dataviz-cursorrules-prompt-file
40.6kCursor rules for Semiotic data visualization library with 30+ chart types, MCP server, and AI-assisted chart generation.
Agent-Reach
71.9kGive your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.
ruflo
67.9k🌊 The original agent meta-harness. Deploy intelligent multi-player swarms, coordinate autonomous workflows, and build conversational AI systems. Features adaptive memory, self-learning intelligence, RAG integration, and native Claude Code / Codex / Hermes and many more Integrated
