Open MCP Server

How to use

Connect to your MCP client and start issuing requests to manage prompts, skills, and workflows. You can run standard actions like starting the local server, executing predefined workflows, or rendering prompts with your chosen variables. The server is designed to work in conjunction with a client that can issue tool calls, pass parameters, and handle results from AI summarization and other AI-enhanced features.

How to install

Prerequisites you need on your machine: Node.js and npm installed. You may also need an API key for AI summarization if you plan to use that feature.

Install dependencies and build the project, then start the server.

Configuration and usage notes

Set the API key for AI summarization features if you plan to use those capabilities. The key is provided as an environment variable named GOOGLE_GENERATIVE_AI_API_KEY.

Run the server using the provided startup command in your environment. The recommended local runtime is to execute the Node.js entry point with your workspace path, which points to the compiled server script.

Project configuration example

{
  "mcpServers": {
    "open_mcp": {
      "type": "stdio",
      "name": "open_mcp",
      "command": "node",
      "args": ["/Users/sdluffy/conductor/workspaces/playground/san-jose/open-mcp/dist/index.js"]
    }
  }
}

Security considerations

The server validates requested paths against an allowed directory whitelist, prevents command injection, and enforces timeout protections on HTTP requests and directory traversal safeguards.

Notes on project structure and tools

The MCP server provides a range of tools to manage prompts, skills, and workflows, including listing, searching, rendering, validating, and executing. It also exposes a set of core utilities for file access, web interactions, version control, system information, and AI summarization.

Troubleshooting

If you encounter issues starting the server, verify that Node.js and npm are installed, ensure the environment variable for the API key is set if you use AI features, and confirm the path to the compiled index script is correct in the startup command.

Examples and common workflows

- Start a daily briefing workflow to generate a productivity summary - Execute a project status workflow to gather updates from connected tools - Render a productivity prompt with your current tasks and priorities

Project structure overview

The server is organized into core engines for prompts, skills, and workflows, plus a registry of tools that handle file operations, web scraping, git operations, and AI summarization.

list_prompts

List all prompts with optional category filtering.

search_prompts

Search prompts by a query string.

get_prompt

Retrieve a prompt template by id.

render_prompt

Render a prompt with given parameters.

validate_prompt

Validate prompt parameters before rendering.

get_prompt_categories

List all prompt categories.

summarize_document

Read and summarize a document.

analyze_text

Read and analyze text.

setup_project

Initialize a new project.

daily_briefing

Get a daily productivity briefing.

project_status

Get a project status report.

read_file

Read a file from the filesystem.

write_file

Write content to a file.

list_directory

List files in a directory.

search_files

Search files by content or name.

fetch_url

Fetch content from a URL.

scrape_html

Scrape HTML content from a webpage.

git_status

Show git repository status.

git_log

Show git commit history.

git_diff

Show git diffs.

system_info

Provide system information.

get_time

Get the current time.

summarize

General AI summarize function.