Text Generator Plugin

https://docs.text-gen.com/
Admin
Text Generator is an open-source AI Assistant Tool that integrates generative artificial...
Tokens:23,884
Snippets:312
Trust Score:8.6
Update:8 months ago
Show doc for...
Tokens:
Raw
### Run Helper Function - Block Mode Examples

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/run+helper+function

Demonstrates the block mode 'run' command with examples such as passing 'selection' or direct strings as input, and using titles for brainstorming templates.

```text-generator-template
{{#run "default/summarize" "summary"}}
    {{selection}}
{{/run}}

```

```text-generator-template
{{#run "default/summarize" "summary"}}
text to be summarized
{{/run}}

```

```text-generator-template
{{#run "local/brainstorm" "ideas" "idea"}}
    {{title}}
{{/run}}

```

--------------------------------

### Request Headers Example for Huggingface

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/Template+Metadata

Sends additional request metadata, often used for authentication. This example demonstrates setting the 'Authorization' header with a Bearer token for the Huggingface provider.

```json
{"Authorization": "Bearer {{keys.hf}}"}
```

--------------------------------

### Example Output Path

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/Template+Metadata

Defines the path to retrieve the result from the response when using the template. This path can be customized to extract specific data from the response, such as an image URL.

```markdown
![]({{requestResults.data.0.url}})
```

--------------------------------

### Run Helper Function - Inline Mode Examples

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/run+helper+function

Illustrates various use cases of the inline 'run' command, including passing 'selection' as input, passing direct strings, and using titles as ideas for other templates.

```text-generator-template
{{run "local/summarize" "summary" selection}}

```

```text-generator-template
{{run "local/summarize" "summary" "text to be summarized"}}

```

```text-generator-template
{{run "local/brainstorm" "ideas" title "idea"}}

```

```text-generator-template
brainstorm ideas about {{idea}}

```

--------------------------------

### CopyAI Configuration Example

Source: https://docs.text-gen.com/_notes/old/general/copyai

Example of frontmatter configuration for the CopyAI tool. This snippet shows how to set the title and tags for a note related to CopyAI.

```yaml
title: copyai
tags: 
layout: note 

```

--------------------------------

### Text Generation with @vanakat/plugin-api

Source: https://docs.text-gen.com/_notes/6-+Additional+resources/Use+Text+Generator+into+another+plugins

This section details how to install and use the Text Generator plugin's API to generate text. It includes installation instructions, usage examples, and a complete code example.

```APIDOC
## Text Generation with `@vanakat/plugin-api`

### Description
This API allows you to programmatically generate text using the Text Generator plugin.

### Method
`gen(prompt)`

### Parameters
#### Path Parameters
None

#### Query Parameters
None

#### Request Body
None

### Request Example
```javascript
// Import and initialize the API
const pluginApi = require('@vanakat/plugin-api');
const tg = pluginApi('tg');

// Generate text
async function generateText(prompt) {
  return await tg.gen(prompt);
}

// Use the function to generate text
generateText("Your prompt here").then(console.log).catch(console.error);
```

### Response
#### Success Response (200)
- **text** (string) - The generated text based on the provided prompt.

#### Response Example
```json
{
  "generated_text": "This is the text generated by the AI."
}
```

#### Error Response
- **error** (string) - A message describing the error that occurred during text generation.
```

--------------------------------

### Request Body Example for DALL-E-2

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/Template+Metadata

Provides data for the API request, with the format varying by provider. This example shows the body for the DALL-E-2 provider, including parameters like 'n', 'size', and 'prompt'.

```json
{"n": 1, "size": "1024x1024", "prompt": "{{escp prompt}}"}
```

--------------------------------

### Example: Text Generation Workflow (JavaScript)

Source: https://docs.text-gen.com/_notes/6-+Additional+resources/Use+Text+Generator+into+another+plugins

A complete example demonstrating how to import, initialize the Text Generator API, generate text with a prompt, and log the result or any errors. This showcases a typical asynchronous usage pattern.

```javascript
// Import and initialize the API
const pluginApi = require('@vanakat/plugin-api');
const tg = pluginApi('tg');

// Generate text
(async () => {
  try {
    const text = await tg.gen("Example prompt");
    console.log(text);
  } catch (error) {
    console.error(error);
  }
})();
```

--------------------------------

### Template Metadata Example

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/Template+Metadata

This snippet shows an example of the YAML front matter used for template metadata in Text Generator. It includes fields like title, tags, and layout.

```yaml
title: Template Metadata
tags: 
layout: note
```

--------------------------------

### Frontmatter Configuration Example

Source: https://docs.text-gen.com/_notes/old/general/Considered+Context

This example shows how to configure frontmatter variables for the Text Generator Plugin. These variables can be included as context for text generation, provided they do not conflict with plugin-reserved YAML keywords.

```yaml
---
title: Considered Context
tags: doc
layout: note 
---
```

--------------------------------

### Install @vanakat/plugin-api

Source: https://docs.text-gen.com/_notes/6-+Additional+resources/Use+Text+Generator+into+another+plugins

Installs the @vanakat/plugin-api package using npm. This package provides the necessary tools to interact with the Text Generator plugin.

```bash
npm install @vanakat/plugin-api
```

--------------------------------

### Interact with Other Plugins via PluginApi

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/Javascript+Templates

This example demonstrates how to interact with other installed Obsidian plugins, such as Text To Speech, using the `pluginApi` function within a Text Generator Javascript script. Ensure the target plugin is installed and script execution is enabled.

```html
{{#script}}
const tts = pluginApi('tts');
await tts.say("This is a test"); // Language is optional, use an ISO 639-1 code
{{/script}}
```

--------------------------------

### Frontmatter Configuration Example

Source: https://docs.text-gen.com/_notes/old/general/Examples+of+frontmatter+configurations

This snippet demonstrates a basic frontmatter configuration for a note within the Text Generator plugin. It includes essential metadata like title and tags, and specifies the layout.

```yaml
title: Examples of frontmatter configurations
tags: 
layout: note 
```

--------------------------------

### Display Notice Message to User

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/Handlebar+Templates

The `notice` helper displays a non-critical informational message to the user. This can be used to provide feedback or suggestions without interrupting the user's workflow.

```handlebars
{{notice "Working on something..."}}
```

--------------------------------

### Configure Custom Provider Headers for Hugging Face

Source: https://docs.text-gen.com/_notes/old/templates/draft+metadata

Example of custom headers for the Hugging Face provider, including authorization token.

```json
headers: '{ "Authorization": "Bearer {{keys.hf}}" }'

```

--------------------------------

### Template Metadata Fields

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/Template+Metadata

This section details the various fields available within the Template File Metadata for the Text Generator plugin, outlining their types, purposes, and providing examples.

```APIDOC
## Template Metadata

### Description
The Template File Metadata is a crucial component of any template in Text Generator. It serves as the header information that provides essential details about the template itself. Properly documenting this metadata ensures clarity, usability, and consistency in your template management. This documentation will guide you through the various fields within the Template File Metadata and how to use them effectively.

### Fields in Template File Metadata

| Field             | Type                  | Purpose                                                              | Example                  |
|-------------------|-----------------------|----------------------------------------------------------------------|--------------------------|
| `promptId`        | String                | Unique identifier for each prompt. Must match the name of the template file. |                          |
| `name`            | String                | Concise and descriptive name for the prompt.                         |                          |
| `description`     | String                | Explains the purpose of the prompt.                                  |                          |
| `author`          | String                | Attributes the prompt to its creator.                                |                          |
| `required_values` | Array of Strings      | Lists values necessary for template function.                        |                          |
| `tags`            | Array of Strings      | Relevant keywords for filtering prompts.                             |                          |
| `version`         | String (Semantic Ver.)| Indicates prompt version.                                            |                          |
| `commands`        | Array of Strings      | Specifies actions associated with the template.                      | `generate`, `insert&create` |
| `mode`            | String                | Specifies the desired mode for completions.                          | `insert`, `replace`      |
| `system and messages` | Object              | Configures system behavior and define messages.                      |                          |
| `max_tokens`      | Number                | Maximum tokens considered in completion.                             |                          |
| `temperature`     | Number                | Indicates randomness of model output.                                |                          |
| `stream`          | Boolean               | Controls streaming mode operation.                                   | `true`, `false`          |
| `provider`        | String                | Service/platform used for processing.                                | "openAIChat", "ollama" |
| `model`           | String                | Name of the model for selected provider.                             |                          |
| `disableProvider` | Boolean               | Allows disabling of template execution (disabled by default).        | `true`, `false`          |
| `endpoint`        | String                | Specific endpoint for template use.                                  |                          |
| `output`          | String                | Path to retrieve result from response.                               | `![]({{requestResults.data.0.url}})` |
| `body`            | String (Object)       | Data for the API request.                                            | `{"n": 1, "prompt": "{{escp prompt}}"}` |
| `headers`         | String (Object)       | Sends additional request metadata.                                   | `{"Authorization": "Bearer {{keys.hf}}"}` |
| `chain.loader`    | String                | Loader function to initialize chain process.                         | `loadSummarizationChain` |
| `chain.type`      | "map_reduce" | "refine" | "staff" | Type of langchain chains to utilize.                               |                          |
| `chain.verbose`   | Boolean               | Provides detailed logging for chain process.                         |                          |
| `splitter.chunkSize` | Number             | Size of each chunk for input data splitting.                         |                          |
| `splitter.chunkOverlap` | Number          | Overlapping elements between consecutive chunks.                     |                          |
| `strict`          | Boolean               | Forces the user to fill all the Variables before running the template. |                          |
| `viewTypes`       | Array of Strings      | Tells Obsidan what files are supported (markdown, excalidraw, canvas) |                          |

### `promptId`
  * **Type** : String
  * **Purpose** : The `promptId` field serves as a unique identifier for each prompt in your UI. It distinguishes one template from another, helping users quickly locate and select the desired prompt. It must be the same as the name of the template file.

### `name`
  * **Type** : String
  * **Purpose** : The `name` field provides a concise and descriptive name for the prompt. It should be short yet precise, enabling users to identify the prompt easily.

### `description`
  * **Type** : String
  * **Purpose** : The `description` field explains the purpose of the prompt and how users should utilize it. It is essential to keep this description simple, concise, and user-friendly.

### `author`
  * **Type** : String
  * **Purpose** : The `author` field attributes the prompt to its creator. It helps give credit to the person responsible for creating the template.

### `required_values`
  * **Type** : Array of Strings
  * **Purpose** : The `required_values` field lists the values that are necessary for the template to function correctly. It informs users of the specific data inputs or parameters they need to provide when using the template.

### `tags`
  * **Type** : Array of Strings
  * **Purpose** : The `tags` field includes relevant keywords or labels that describe the prompt. These tags are used for filtering prompts, making it easier for users to find templates that match their needs.
```

--------------------------------

### Fetch Current Date and Time

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/Handlebar+Templates

The `date` helper retrieves the current date and time. This is useful for timestamping operations or any feature requiring current date information.

```handlebars
{{date}}
```

--------------------------------

### Handlebars Template Operation Helper

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/Handlebar+Templates

The `run` helper allows executing another template within the current context, promoting reusability and modular design for complex template structures.

```Handlebars
{{run "templateName"}}
```

--------------------------------

### Initialize Text Generator API (JavaScript)

Source: https://docs.text-gen.com/_notes/6-+Additional+resources/Use+Text+Generator+into+another+plugins

Initializes the Text Generator API by importing the plugin API and creating an instance for 'tg'. This setup is required before making any text generation calls.

```javascript
const pluginApi = require('@vanakat/plugin-api');
const tg = pluginApi('tg');
```

--------------------------------

### Select Random File Based on Pattern and Length

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/Handlebar+Templates

The `getRandomFile` helper selects a random file that matches a given pattern in its path and falls within a specified content length range (minLength and maxLength).

```handlebars
{{getRandomFile "input/" 100 1500}}
```

```handlebars
{{getRandomFile pattern minLength maxLength}}
```

--------------------------------

### Iterating Through {{highlights}}

Source: https://docs.text-gen.com/_notes/3-+Templates/01+Understanding+Context

The {{#each highlights}} helper iterates over text segments marked with double equal signs (==example==). It provides a way to selectively process highlighted content within a note.

```text-generator-template
{{#each highlights}}
* {{this}}
{{/each}}
```

--------------------------------

### Configure Custom Provider Body for DALL-E 2

Source: https://docs.text-gen.com/_notes/old/templates/draft+metadata

Example of a custom request body for the DALL-E 2 provider, including parameters like number of images, size, and prompt.

```json
body: "{\"n\": 1, \"size\": \"1024x1024\", \"prompt\": \"{{escp prompt}}\"}"

```

--------------------------------

### Run Helper Function - Block Mode Additional Context

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/run+helper+function

Shows how to pass additional context or variables within the block mode 'run' command, emphasizing the need for JSON-serializable data structures.

```text-generator-template
{{#run "default/summarize" "summary"}}
{
  "selection": "{{escp2 title}}",
  ... other config
}
{{/run}}

```

--------------------------------

### Run Helper Function - Block Mode Syntax

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/run+helper+function

Presents the syntax for the block mode 'run' command in Text Generator templates. This mode allows for passing content within the block as input to the executed template.

```text-generator-template
{{#run "packageId/promptId" "output" "target"}}
    {{input}}
{{/run}}

```

--------------------------------

### Log Message to Console

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/Handlebar+Templates

The `log` helper outputs a message to the console. It is primarily used for debugging purposes or to inform the user about background processes within the template.

```handlebars
{{log "Working on something..."}}
```

--------------------------------

### Handlebars File Operation Helpers

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/Handlebar+Templates

Helpers for performing file operations within Handlebars templates. The `write` helper writes content to a specified file, and the `read` helper retrieves content from a file.

```Handlebars
{{#write "readme.md"}}
... content of readme.md file in the document ...
{{/write}}
```

```Handlebars
{{read "readme.md"}}
```

--------------------------------

### Adding Starred Blocks for Context

Source: https://docs.text-gen.com/_notes/old/general/Considered+Context

This example demonstrates how to include specific sections of a long article as context by adding an asterisk to the heading. This requires enabling the option in the plugin's settings.

```markdown
# Title*
This is a title
# Introduction
write introduction
```

--------------------------------

### Iterating Through Linked Mentions

Source: https://docs.text-gen.com/_notes/3-+Templates/01+Understanding+Context

The {{#each mentions.linked}} helper iterates through notes that are directly linked to the current note within the vault. It facilitates working with explicitly connected notes.

```text-generator-template
{{#each mentions.linked}}
* {{this.results}}
{{/each}}
```

--------------------------------

### Set Completion Mode

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/Template+Metadata

Specifies the desired mode for completions when using the template. Available modes include 'insert', 'replace', and 'rename'. This field offers flexibility in how completions are integrated into content.

```yaml
mode: insert
```

--------------------------------

### Using {{clipboard}} Context Variable

Source: https://docs.text-gen.com/_notes/3-+Templates/01+Understanding+Context

The {{clipboard}} context variable represents the current content that has been copied to the system clipboard. It enables templates to interact with and insert clipboard data.

```text-generator-template
{{clipboard}}
```

--------------------------------

### Template File Metadata (PromptInfo) YAML Example

Source: https://docs.text-gen.com/_notes/old/templates/Template+File+Metadata

This YAML snippet shows the 'PromptInfo' section, which contains essential metadata for a prompt template. It includes fields for prompt identification, naming, description, author, required values, tags, version, and commands.

```yaml
---
PromptInfo:
 promptId: getParagraph
 name: ✍️ Write paragraphs 
 description: select a content contiens items, a paragraph for each item will be generated. 
 required_values: title, outline
 author: Noureddine
 tags: writing
 version: 0.0.1
 commands:
 - generate
---

```

--------------------------------

### Run Helper Function - Inline Mode Syntax

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/run+helper+function

Defines the syntax for the inline 'run' command in Text Generator templates. It allows executing a template and optionally storing its output and passing input.

```text-generator-template
{{run "packageId/promptId" "output" "input" "target"}}
```

--------------------------------

### Template File Metadata YAML Example

Source: https://docs.text-gen.com/_notes/old/templates/Template+File+Metadata

This YAML snippet demonstrates the structure for defining metadata for a template file. It includes fields like title, tags, and layout, which are common in static site generators or documentation systems.

```yaml
title: Template File Metadata
tags: 
layout: note 

```

--------------------------------

### Iterating Through {{selections}}

Source: https://docs.text-gen.com/_notes/3-+Templates/01+Understanding+Context

The {{#each selections}} helper iterates over all text segments selected by the user, particularly useful for multiple selections made with the Alt key. It allows systematic processing of each selected piece of text.

```text-generator-template
{{#each selections}}
* {{this}}
{{/each}}
```

--------------------------------

### Using {{content}} Context Variable

Source: https://docs.text-gen.com/_notes/3-+Templates/01+Understanding+Context

The {{content}} context variable provides access to the entire text content of the current note. This is useful for operations that require the full note body.

```text-generator-template
{{content}}
```

--------------------------------

### Configure System and Messages for LLM

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/Template+Metadata

Configures the system's behavior and defines messages for interactions with the LLM. The 'system' field is a string describing the system's role, and 'messages' is an array of alternating user and assistant messages to define conversation flow.

```yaml
system: <Your system description>
messages:
- <user Message>
- <assistant Message>
- <user Message>
- <assistant Message>
```

--------------------------------

### Using {{title}} Context Variable

Source: https://docs.text-gen.com/_notes/3-+Templates/01+Understanding+Context

The {{title}} context variable represents the title of the current note. It can be used to dynamically reference or manipulate the note's title within templates.

```text-generator-template
{{title}}
```

--------------------------------

### Split Text using Langchain in Javascript

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/Javascript+Templates

This Javascript example utilizes Langchain's `RecursiveCharacterTextSplitter` to split text extracted from a YouTube video into manageable chunks. It requires the `extract` function and Langchain's splitter. Ensure script execution is enabled.

```html
{{#script}}
	const youtubeScript = await extract("yt", "https://www.youtube.com/watch?v=tNAsLbGdM6A");
	
	const splitter = new RecursiveCharacterTextSplitter({
	  chunkSize: 4000,
	  chunkOverlap: 1,
	});
	
	const output = await splitter.createDocuments([youtubeScript]);
	
	return output.join("***/n");
{{/script}}
```

--------------------------------

### Select Next/Previous Word Relative to Cursor in Text Generation

Source: https://docs.text-gen.com/_notes/3-+Templates/01+Understanding+Context

The `{{NextWord}}` and `{{PreviousWord}}` context variables select the next or previous word relative to the cursor's position, respectively. These facilitate word-by-word navigation and editing.

```handlebars
{{NextWord}}
{{PreviousWord}}
```

--------------------------------

### Iterate YAML Metadata in Text Generation

Source: https://docs.text-gen.com/_notes/3-+Templates/01+Understanding+Context

The `{{#each yaml}}` helper iterates through the top-level metadata (Object) of a note. It provides access to keys (`{{@key}}`) and values (`{{this}}`) of the metadata. For direct access to a specific variable, use `{{yaml.variable}}`.

```handlebars
{{#each yaml}}
{{@key}}: {{this}}
{{/each}}

{{yaml.variable}}
```

--------------------------------

### Iterating Through Unlinked Mentions

Source: https://docs.text-gen.com/_notes/3-+Templates/01+Understanding+Context

The {{#each mentions.unlinked}} helper iterates through notes that are referenced in the vault without a direct wiki-link (e.g., '...note...'). It helps manage and process unlinked references to the current note.

```text-generator-template
{{#each mentions.unlinked}}
* {{this.results}}
{{/each}}
```

--------------------------------

### Use substring function in templates

Source: https://docs.text-gen.com/_notes/old/commands/substring

The 'substring' function allows extraction of a portion of text. It takes the variable name, a start position, and an end position as arguments. For example, to get the first 100 characters of 'selection', use '{{{substring selection 0 99}}}'.

```text/template
{{{substring var startPosition endPosition}}}
```

```text/template
{{{substring selection 0 99}}}
```

--------------------------------

### Choose a model Configuration (YAML)

Source: https://docs.text-gen.com/_notes/old/general/Choose+a+model

This snippet demonstrates how to configure the 'Choose a model' setting using YAML frontmatter. It specifies the title, tags, and layout for the note. The model selection can be overridden by frontmatter settings and supports the latest ChatGPT models. The list of available OpenAI models can be updated in the plugin settings.

```yaml
title: Choose a model
tags: 
layout: note 
```

--------------------------------

### Frontmatter Configuration

Source: https://docs.text-gen.com/_notes/old/general/Frontmatter+Configuration

Explains how to configure the Text Generator Plugin using frontmatter, including options for appending parameters, context, and output handling.

```APIDOC
## Frontmatter Configuration

### Description
Frontmatter configuration allows for flexible setup of the Text Generator Plugin, compatible with services like OpenAI and HuggingFace. It defines how requests are structured and processed.

### Method
N/A (Configuration)

### Endpoint
N/A (Configuration)

### Parameters
#### Path Parameters
N/A

#### Query Parameters
N/A

#### Request Body
N/A (Configuration is done via frontmatter)

### Request Example
```yaml
title: Frontmatter Configuration
tags: doc
layout: note
```

## YAML configuration 
Frontmatter configuration is very flexible and can work with many services such as OpenAI and HuggingFace.
## The structure of the frontmatter 
```yaml
---
config:
 append:
  bodyParams: Boolean
  reqParams: Boolean
 context: String
 output: String
bodyParams:
 parameter: value
reqParams:
 parameter: value
---
```

### Configuration Options:
*   **config**
    *   **append**
        *   **bodyParams** : Append or not body form-data parameters to the default configuration (true/false). Default is `true`.
        *   **reqParams** : Append or not request parameters to the default configuration (true/false). Default is `true`.
    *   **context**
        *   The variable that will include the considered content in `bodyParams`. Default is `"prompt"`.
    *   **output**
        *   Request results are in the variable `requestResults`. Requires Javascript code to extract results.
  *   **bodyParams**
      *   Includes body form-data parameters. Refer to the specific API documentation (e.g., OpenAI Documentation) for available variables.
      *   Examples:
          *   **n** : Number of generations
          *   **max_tokens** : Maximum number of tokens per generation.
  *   **reqParams**
      *   **url** : Base URL for request construction.

### See Examples of frontmatter configurations.
[Link to examples]
```

--------------------------------

### Handlebars Variable Management Helpers

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/Handlebar+Templates

Helpers for managing variables within Handlebars templates. The `set` helper stores values in variables, accessible via the `vars` object, while the `get` helper retrieves these stored values.

```Handlebars
{{set "var1" selection}}
```

```Handlebars
{{#set "var1"}}
 {{selection}}
{{/set}}
```

```Handlebars
{{get "var1"}}
or
{{vars.var1}}
```

--------------------------------

### Run Another Template using Run Helper Function

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/Javascript+Templates

This Javascript example shows how to execute another Text Generator template (e.g., 'default/summary') from within a script, passing data as context. It first extracts content using `extract` and then uses `run` to process it. Script execution must be enabled.

```html
{{#script}}
	const youtubeScript = await extract("youtube", "https://www.youtube.com/watch?v=tNAsLbGdM6A");
	
	const summary = await run("default/summary", {tg_selection: youtubeScript});
	
	return summary;
{{/script}}
```

--------------------------------

### Configure System and Messages for Interactions

Source: https://docs.text-gen.com/_notes/old/templates/draft+metadata

Demonstrates setting up system prompts and message arrays for structured interactions with the language model, useful for tasks like translation or correction.

```yaml
system: "You are an English translator, spelling corrector, and improver. Respond in corrected and improved English."
messages:
  - "User: Hello, how are you?"
  - "Assistant: Hello! I am doing well, thank you for asking."
  - "User: I need help with this sentence."
  - "Assistant: Please provide the sentence, and I will do my best to assist you."

```

--------------------------------

### Create New File From Template - Text Generator

Source: https://docs.text-gen.com/_notes/2-+Options/Commands/Generate+and+Create+a+New+File+From+Template

This command generates a new file by utilizing a specified template and processing a prompt. The output of the prompt is then used to populate the new document.

```markdown
title: Generate and Create a New File From Template
tags: 
layout: note 

```

--------------------------------

### Configure Model Parameters and Streaming

Source: https://docs.text-gen.com/_notes/old/templates/draft+metadata

Shows how to set parameters like maximum tokens and temperature for controlling the language model's output, and enable/disable streaming mode.

```yaml
# Model parameters
max_tokens: 150
temperature: 0.7

# Streaming mode
stream: true

```

--------------------------------

### Trim Whitespace from String

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/Handlebar+Templates

The `trim` helper removes leading and trailing whitespace from a string. This is useful for cleaning user input or preparing text for output.

```handlebars
{{trim "    hello world    "}}
```

--------------------------------

### Handlebars Information Extraction Helper

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/Handlebar+Templates

The `extract` helper is used to obtain information from various sources, enhancing the template's content processing and analysis capabilities.

```Handlebars
{{extract "source"}}
```

--------------------------------

### Basic Content Extraction with 'extract' (Text Generator)

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/extract+helper+function

Demonstrates the basic usage of the 'extract' command to retrieve content from various sources. It supports different extractor types like 'web', 'youtube', 'pdf', 'image', and 'audio', taking the source as an argument.

```text-generator
{{extract "extractorType" source}}
```

```text-generator
{{extract "web" weburl}}
{{extract "web" "https://example.com"}}
```

```text-generator
{{extract "youtube" ytUrl}}
{{extract "pdf" pdfPath}}
{{extract "image" imgPath}}
{{extract "audio" audioPath}}
```

--------------------------------

### Filter Unique Elements from Array

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/Handlebar+Templates

The `unique` helper filters an array to return only unique elements, removing any duplicates. It takes an array as input and returns a new array with distinct values.

```handlebars
{{unique ["apple", "banana", "apple", "orange", "banana"]}}
```

--------------------------------

### Configure Provider and Endpoint Settings

Source: https://docs.text-gen.com/_notes/old/templates/draft+metadata

Details how to specify the AI provider, custom endpoints, and disable providers if necessary.

```yaml
provider: "openAIChat"
# endpoint: "https://api.example.com/v1"
disableProvider: false

```

--------------------------------

### Iterating Through {{children}}

Source: https://docs.text-gen.com/_notes/3-+Templates/01+Understanding+Context

The {{#each children}} context variable iterates over notes or sub-notes that are cited or related to the current note. It allows access to and display of information from related notes.

```text-generator-template
{{#each children}}
# {{this.title}}
{{this.content}}
{{/each}}
```

--------------------------------

### Excalidraw Plugin Parsing

Source: https://docs.text-gen.com/assets/images/childrenNotes

Demonstrates how the Excalidraw plugin parses text elements and templates. It shows the structure for referencing child notes and iterating through them.

```markdown
excalidraw-plugin: parsed
tags: [excalidraw]

⚠ Switch to EXCALIDRAW VIEW in the MORE OPTIONS menu of this document. ⚠
# Text Elements 
CH2 
Template 
[[ 
[[ 
CH1 
[[ 
[[ 
CH2 
CH1 
Note 
{{children[1]}} 
{{children[0]}} 
{{ each children}}  
{{this}}  
{{/each}}
```

--------------------------------

### Display Error Message and Halt Execution

Source: https://docs.text-gen.com/_notes/3-+Templates/subpages/Handlebar+Templates

The `error` helper displays a critical error message to the user and halts the template's execution. It is used when essential conditions are not met and the template cannot proceed.

```handlebars
{{error "Sorry, this template won't work without a note"}}
```

--------------------------------

### Define Completion Modes in Configuration

Source: https://docs.text-gen.com/_notes/old/templates/draft+metadata

Illustrates how to configure the desired mode for text completions (insert, replace, rename) within the plugin's settings.

```yaml
# Configuration for completion modes
# Available modes: insert, replace, rename
mode: insert

```

--------------------------------

### Using {{selection}} Context Variable

Source: https://docs.text-gen.com/_notes/3-+Templates/01+Understanding+Context

The {{selection}} context variable refers to the text currently selected by the user. It allows templates to operate on specific highlighted portions of a note.

```text-generator-template
{{selection}}
```

--------------------------------

### Manifest File Configuration for Template Packages

Source: https://docs.text-gen.com/_notes/old/templates/Share+a+New+Template+Package

Defines the structure and required fields for a manifest file used in Text Generator plugin template packages. This file includes metadata such as package ID, name, version, description, author, and repository information.

```json
{
	"packageId": "default", // package id has to be unique from the other packages
	"name": "Default Prompts Package", // name of the package
	"version": "0.0.6", // version of the package if you're planning to update it later
	"minTextGeneratorVersion": "0.1.0", // minimum version of text generator that can work with it (use your current version)
	"description": "...", // description of the package
	"author": "...", // author name
	"tags" : "writing, brainstorming", // tags for the package
	"authorUrl": "...", // author page, or funding page
	"repo":"username/repo" // get it from the repo url
}
```

--------------------------------

### Select Current Paragraph at Cursor in Text Generation

Source: https://docs.text-gen.com/_notes/3-+Templates/01+Understanding+Context

The `{{cursorParagraph}}` context variable selects the entire paragraph where the cursor is currently located. This is helpful for operations focused on paragraph-level editing, formatting, or analysis.

```handlebars
{{cursorParagraph}}
```
Tokens:
Raw