mcp -project-standards

This MCP server provides a centralized set of project standards management tools for AI-assisted development. It supports organizing, validating, and applying unified development standards across multiple machines and teams. The server exposes tools that help you manage project information, analyze project structure, define API and development standards, and handle configuration in a JSON-based format. You can use the server to generate and enforce standard rules, inspect directories for standard-compliant structures, and download necessary resources directly into your project path. Typical workflows include bootstrapping a project with a default configuration, generating or updating standard rules, and validating code artifacts against the defined standards. If you run the server with the recommended npx-based startup, you’ll have access to a suite of tools like list_directory, generate_rules, generate_cursorrules, and download_file, all designed to keep your development environment aligned with your project standards. When combined with environment-based project isolation (via TOOL_PREFIX) the MCP server can manage multiple projects with minimal configuration changes across machines.

Prerequisites:

• Node.js (LTS version) and npm installed on your machine.
• Internet access to install the MCP package.

Step-by-step:

1. Install Node.js and npm from https://nodejs.org/ (or via your system package manager).

2. Install the MCP package globally (recommended) so you can run it from any directory:

   npm install -g @liangshanli/mcp

3. Run the MCP server using npx (as configured in mcp_config):

   npx -y @liangshanli/mcp

4. If you use a custom project path or configuration directory, set environment variables before starting:

   • On Unix-like systems: export PROJECT_PATH=/path/to/your/project export CONFIG_DIR=/path/to/config
   • On Windows (PowerShell): $env:PROJECT_PATH = 'C:\path\to\your\project' $env:CONFIG_DIR = 'C:\path\to\config'

5. Verify the MCP server is running and accessible via the expected interface (console output will indicate readiness).

Optional: You can also run via npx with explicit package if you prefer not to install globally:

npx -y @liangshanli/mcp

Tips and notes:

• Environment variables like PROJECT_PATH and CONFIG_DIR control where the server resolves relative paths and stores/reads configuration. Set TOOL_PREFIX and PROJECT_NAME if you need multi-project isolation (as described in the v3.0.0 breaking changes) to route tools to the correct project context.
• The server supports a JSON-based configuration and auto-creates default configurations on first run; ensure your CONFIG_DIR is writable.
• If you enable project isolation, you can switch project contexts by changing environment variables rather than code changes, facilitating smoother multi-machine collaboration.
• The tool suite includes download_file, list_directory, generate_rules, and generate_cursorrules. Use these to fetch resources, inspect project structure, and generate standardized rules for API design, development practices, and database conventions.
• If you encounter issues related to path resolution, review the CONFIG_DIR and TOOL_PREFIX settings described in the v3.0.0 release notes to ensure proper per-project isolation.