cursorrules

A project rule for an MCP server that connects an AI client to a Minecraft bot. MCP is a framework for exposing actions and read-only data, while JSON-RPC is the message format it handles behind the scenes.

In plain words
What is it for?
Use it when adding or changing Minecraft bot actions, such as digging or navigating, and read-only data such as inventory access.
Why use it?
It keeps the implementation focused on Minecraft and Mineflayer behavior instead of duplicating the framework's communication code.

Cursor rule for Cursor

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add rules/gerred/mcpmc/cursorrules
Clone the repo
git clone --depth 1 https://github.com/gerred/mcpmc

Made for: Cursor.

Per session 2,356 This file is loaded in full into every session.
When invoked 2,356 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.02356 $0.02356
Opus 5 $0.01178 $0.01178
Sonnet 5 $0.00471 $0.00471
Haiku 4.5 $0.00236 $0.00236

Measured 2d ago against content hash 371bf8aefe1a, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

cursorrules scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.cursorrules · 276 lines

How it starts

The opening of the file, as written. The whole thing — 276 lines — stays where its author put it; the contents beside it link to each section on GitHub.

.cursorrules

⚠️ IMPORTANT: JSON-RPC Warning

If you find yourself implementing JSON-RPC directly (e.g., writing JSON messages, handling protocol-level details, or dealing with stdio), STOP! You are going in the wrong direction. The MCP framework handles all protocol details. Your job is to:

  1. Implement the actual Minecraft/Mineflayer functionality
  2. Use the provided bot methods and APIs
  3. Let the framework handle all communication

Never:

  • Write JSON-RPC messages directly
  • Handle stdio yourself
  • Implement protocol-level error codes
  • Create custom notification systems

Overview

This project uses the Model Context Protocol (MCP) to bridge interactions between a Minecraft bot (powered by Mineflayer) and an LLM-based client.

The essential flow is:

  1. The server starts up ("MinecraftServer") and connects to a Minecraft server automatically (via the Mineflayer bot).
  2. The MCP server is exposed through standard JSON-RPC over stdio.
  3. MCP "tools" correspond to actionable commands in Minecraft (e.g., "dig_area", "navigate_to", etc.).
  4. MCP "resources" correspond to read-only data from Minecraft (e.g., "minecraft://inventory").

When an MCP client issues requests, the server routes these to either: • The "toolHandler" (for effectful actions such as "dig_block")
• The "resourceHandler" (for returning game state like position, health, etc.)

MCP Types and Imports

When working with MCP types:

  1. Import types from the correct SDK paths:
    • Transport: "@modelcontextprotocol/sdk/shared/transport.js"
    • JSONRPCMessage and other core types: "@modelcontextprotocol/sdk/types.js"
  2. Always check for optional fields using type guards (e.g., 'id' in message)
  3. Follow existing implementations in example servers when unsure
  4. Never modify working type imports - MCP has specific paths that must be used

Progress Callbacks

For long-running operations like navigation and digging:

  1. Use progress callbacks to report status to MCP clients
  2. Include a progressToken in _meta for tracking
  3. Send notifications via "tool/progress" with:
    • token: unique identifier
    • progress: 0-100 percentage
    • status: "in_progress" or "complete"
    • message: human-readable progress

Read the full file on GitHub · 276 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 276 lines · 2,356 tokens per session scan A 371bf8aefe1a

Subscribe to this mod's changes

cursorrules is a cursor rule published in the GitHub repository gerred/mcpmc (40 stars, last pushed 1y ago), licensed MIT. It adds 2,356 tokens to every session, about $0.0118 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.