playbooks

home / mcp / moltmd mcp server

Moltmd MCP Server

Provides MCP-based access to molt-md, enabling encrypted markdown documents to be read, written, and managed via an LLM-friendly API.

Installation
Add the following to your MCP client configuration file.

Configuration

View docs
{
  "mcpServers": {
    "bndkts-molt-md-mcp": {
      "command": "molt-mcp",
      "args": [],
      "env": {
        "MOLT_API_KEY": "YOUR_API_KEY",
        "MOLT_BASE_URL": "https://api.molt-md.com/api/v1",
        "MOLT_WORKSPACE_ID": "your-workspace-id-here"
      }
    }
  }
}

You can turn your encrypted markdown documents hosted in molt-md into a fully accessible knowledge base for your AI assistant. This MCP server exposes molt-md endpoints as tools, enabling you to read, write, organize, and manage your markdown content through a simple, secure interface.

How to use

You will run the MCP server locally or in your environment and connect your MCP client with an authentication key. Use the available tools to read documents, create or update documents, manage workspaces, and inspect the state of the database. The server enforces access using a dual-key model where read keys allow only read operations and write keys permit read, create, update, and delete actions.

How to install

Prerequisites: you need Python installed to run the MCP server, along with a working internet connection to install dependencies.

uvx --from git+https://github.com/bndkts/molt-md-mcp molt-md-mcp

Additional configuration and usage notes

Install from source for development if you want to modify the server. After installing, you can start the MCP server by invoking the runtime command provided by the installation process.

git clone https://github.com/bndkts/molt-md-mcp.git
cd molt-md-mcp
uv pip install -e .

Configuration and security notes

Environmental keys control access to molt-md through the MCP server. Prepare the keys and, if you want to scope access to a specific workspace, set the workspace context.

Development and run instructions

To run the MCP server locally after installation, start the server process named molt-mcp.

molt-mcp

Available tools

health_check

Verify that the molt-md API is reachable and responsive.

get_metrics

Retrieve database statistics such as document and workspace counts.

read_doc

Read a document's decrypted content. Supports partial fetches with a lines parameter and can return metadata or plain markdown.

read_workspace

Read a workspace's content, including its name and document entries. Can generate previews with a specified number of preview lines.

create_doc

Create a new encrypted document and receive its ID along with write and read keys.

update_doc

Replace a document's entire content, with optimistic locking using an if_match version tag.

append_doc

Append content to the end of a document, with optimistic locking support.

delete_doc

Permanently delete a document.

create_workspace

Create a new workspace to bundle documents and obtain its ID with write and read keys.

update_workspace

Replace a workspace's content (name and entries) with optimistic locking.

delete_workspace

Permanently delete a workspace without removing the documents it references.