Teleprompter

Exposes prompt storage, search, and variable-enabled templates to LLMs via MCP.

typescript

4

GitHub Stars

typescript

Language

3 months ago

First Indexed

3 weeks 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": {
    "raisinbread-teleprompter": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-teleprompter"
      ],
      "env": {
        "PROMPT_STORAGE_PATH": "/path/to/your/prompts-directory"
      }
    }
  }
}

Teleprompter is an MCP server that stores, searches, and exposes prompt templates for reuse with large language models. It provides a streamlined way to manage dynamic prompts with template variables, so you don’t have to rewrite instructions for every interaction. You can configure Teleprompter to run locally as an MCP server and access its prompt tooling through your MCP client ecosystem.

How to use

You connect to Teleprompter through an MCP client to access prompt templates stored in your designated directory. Use template variables like {{name}} to inject dynamic content into prompts. Store prompts as markdown files and leverage full-text search to quickly locate the right prompt for a given task. You invoke a ready-made prompt by its ID, then fill in the required variables before sending it to your LLM.

How to install

Prerequisites: ensure you have Node.js and npm installed on your machine.

Install Node.js and npm if they are not already available on your system.
Prepare the MCP configuration for Teleprompter by using the following JSON snippet. This config runs Teleprompter via npx and points to your prompts directory.

{
  "mcpServers": {
    "teleprompter": {
      "command": "npx",
      "args": ["-y", "mcp-teleprompter"],
      "env": {
        "PROMPT_STORAGE_PATH": "/path/to/your/prompts-directory"
      }
    }
  }
}

Configuration notes

The Teleprompter MCP server relies on a local prompts directory. Set PROMPT_STORAGE_PATH to the absolute path where your prompts (markdown files) are stored. The server exposes prompt tooling via MCP so your LLM client can search, retrieve, and render prompts with dynamic variables.

Usage patterns and examples

Store prompts with descriptive IDs such as spotify-discover or meeting-notes to organize your library. Use {{variable}} placeholders for dynamic content.
Search prompts quickly using the built-in full-text search over your stored markdown files.

Troubleshooting

If the prompts directory is not found or unreachable, verify that PROMPT_STORAGE_PATH points to a valid absolute path and that your user has read/write permissions. Check that Node.js and npm are installed and that the MCP client can reach the Teleprompter server according to your environment.

Available tools

prompt storage manager

Manages the collection of prompts stored as markdown files, including listing, adding, editing, and organizing prompts by ID and category.

prompt search

Performs fast full-text search over stored prompts to quickly locate templates based on keywords or variables.

template rendering with variables

Renders prompts by substituting template variables like {{name}} with user-provided values to produce a concrete prompt for the LLM.

MCP exposure

Exposes prompt tooling to an MCP client so you can reuse and compose prompts across different LLM calls.