promptapi_spec_generator
Generate PromptAPI specifications based on input parameters.
Examples:
regex_generator
Generate a regular expression that matches a set of strings.
Examples:
Example 1:
Input:
{
"strings_to_match": [
"foo",
"bar",
"baz"
]
}Output:
{
"regex": "^(foo|bar|baz)$"
}Example 2:
Input:
{
"regex_description": "A regex that matches emails",
"strings_to_match": [
"w.seagar@gmail.com"
]
}Output:
{
"regex": "^[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\\.[a-zA-Z]{2,4}$"
}sentiment-extractor
Extract the sentiment of the text.
Examples:
Example 1:
Input:
{
"input_text": "The customer service at this store is absolutely fantastic. Every time I visit, I'm greeted with a warm smile and helpful staff who go above and beyond to make sure I find what I need."
}Output:
{
"sentiment": 10,
"reason": "These words and phrases convey a sense of satisfaction and appreciation for the customer service at the store, indicating a positive sentiment towards the experience.",
"words_or_phrases": [
"absolutely fantastic",
"warm smile",
"helpful staff",
"go above and beyond"
]
}transaction-classifier
Classify a transaction type based on the transaction information provided in the input.
Examples:
Example 1:
Input:
{
"transaction_amount": 1602.25,
"transaction_description": "LOAN PAYMENT ANZ TDM 1234567890"
}Output:
{
"transaction_type_primary": "LOAN_PAYMENTS",
"transaction_type_detailed": "LOAN_PAYMENTS_MORTGAGE_PAYMENT",
"confidence_score": 0.85
}