Prompter CLI

Prompter CLI is a command line tool for managing prompts, bots, and embeddings.

Install

npm install -g prompter-cli

Usage

Verify the installation by running prompter --version in your terminal.

prompter --version

Login

Login to your Prompter account by running prompter login in your terminal.

prompter login

Publish a prompt

Publish a prompt by running prompter --prompt=<prompt> --publish in your terminal. Note that the prompt must be a valid Handlebars template. See Handlebars for more information. If no promptId is provided, the file name will be used as the promptId. If the prompt is already published, it will be updated.

 prompter --prompt=demo.hbr --publish

Publish using a different promptId:

   prompter --prompt=demo.hbr --id=demo --publish

Uploading Plain Text Embeddings

Create a text file with extension .txt then follow the instructions below. Embeddings can be uploaded by running prompter --embed=<file> --collection=<collection-name> in your terminal. Embeddings must be a text file, it supports Mark Down and Json.

you can specifiy chunkSize by running prompter --embed=<file> --collection=<collection-name> --chunkSize=<chunkSize> in your terminal. The default chunkSize is 1000.

Uploading Markdown Embeddings

Create a text file with extension .md then follow the instructions below. Embeddings can be uploaded by running prompter --embed=<file> --collection=<collection-name> in your terminal. Embeddings must be a text file with extension .md. Prompter will detect the file extension and upload the embeddings as markdown.

Additional parameters can be specified by running prompter --embed=<file> --splitTables=true in your terminal. In Markdown, tables are represented as text. Prompter will detect tables and split them into separate chunks keeping a copy of the header in each chunk. This is useful for creating a table of contents.

In markdown, the chunk split criteria is praragraphs, each paragraph will be a separate chunk, keeping a breadcrumb of the parent headlines on each chunk.

Querying Embeddings

Execute prompter then select Embeddings from menu.