Prompteka

Local, high-speed access layer to read and write your Prompteka prompt library via MCP clients.
  • typescript

1

GitHub Stars

typescript

Language

4 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs
{
  "mcpServers": {
    "webdevguyrg-prompteka-mcp": {
      "command": "prompteka-mcp",
      "args": [],
      "env": {
        "LOG_LEVEL": "info",
        "PROMPTEKA_DB_PATH": "/custom/path/prompts.db"
      }
    }
  }
}

You can access and manage your Prompteka prompt library directly from your AI assistant using the Prompteka MCP Server. It runs locally and connects to Prompteka’s SQLite database, letting you read, search, create, update, delete, and organize prompts without leaving your automation flow. Changes are written immediately to your local data, providing a fast, secure, and zero-network workflow.

How to use

Use the MCP client in your AI tool to perform read and write operations against your Prompteka library. Typical workflows include reading folders and prompts, creating new prompts in a chosen folder, updating prompt content, moving prompts between folders, and batch organizing your library. After you perform write actions, refresh Prompteka to reflect changes in the UI.

Common usage patterns you can perform with your MCP client include searching for prompts by keyword, selecting a prompt to view or execute, creating folders, moving prompts, and deleting prompts or folders with safety checks. All actions are executed through the MCP server and apply directly to your local Prompteka database.

How to install

Prerequisites you need installed on your Mac include Prompteka, Node.js, and npm. Ensure Prompteka has been opened at least once to initialize its database.

Step 1: Install the MCP server globally

npm install -g prompteka-mcp

Step 2: Configure your MCP client to connect to the local MCP server

{
  "mcpServers": {
    "prompteka": {
      "command": "prompteka-mcp",
      "args": [],
      "env": {
        "LOG_LEVEL": "info"
      }
    }
  }
}

Step 3: (Optional) verify the MCP server binary is accessible

which prompteka-mcp

Configuration

The MCP server connects directly to Prompteka's SQLite database located in your home folder. You can customize environment variables and start options as shown below.

{
  "mcpServers": {
    "prompteka": {
      "command": "prompteka-mcp",
      "args": [],
      "env": {
        "LOG_LEVEL": "info"
      }
    }
  }
}

If you need to point to a different database path for Prompteka, you can set PROMPTEKA_DB_PATH in the environment when starting the MCP server.

Security & privacy

All operations are local and do not leave your machine. The server reads and writes directly to your Prompteka database with WAL mode enabled for safe concurrent access. No telemetry or network calls are performed.

Ensure proper file permissions and backup practices. Regular backups of your Prompteka library are recommended and supported by the Prompteka app.

Troubleshooting

If you see a database not found error, make sure Prompteka has been installed and opened at least once to create the database.

If the MCP server cannot write due to a database lock, retry the operation. The system uses atomic transactions and WAL mode to minimize conflicts.

If changes aren’t visible in Prompteka after MCP writes, refresh the UI in Prompteka (⌘R or Refresh). The MCP server writes immediately, and a UI refresh is needed to display the latest data.

Available tools

list_folders

Read operation that retrieves all folders with metadata including hierarchy and timestamps.

list_prompts

Read operation to fetch prompts from a specific folder with pagination.

get_prompt

Read operation to fetch full details of a single prompt by its ID.

search_prompts

Read operation that performs a full-text search across prompt titles and content.

health_check

Read operation to verify MCP server and database are operational and returns status.

create_prompt

Write operation to create a new prompt in a folder with optional emoji, color, and URL.

update_prompt

Write operation to modify an existing prompt's content, metadata, and location.

delete_prompt

Write operation to delete a prompt with safety idempotent guarantees.

create_folder

Write operation to create a new folder with an optional parent for nesting.

update_folder

Write operation to rename or reorganize a folder.

delete_folder

Write operation to delete a folder, with safety checks and optional recursive deletion.

move_prompt

Write operation to move a prompt to a different folder.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational