Using FluentCLI

The core of fluent is taking the various types of input from a command prompt and allowing those to be combined in interesting ways to send to an AI LLM for processing.

---

A most basic FluentCLI command example is:

This command calls the OpusChain (we don't know if it's Langflow or Flowise from here) and sends the request 'This is a test' to the chatflow.

A few moments later, the response is returned to the terminal as text.

fluent OpusChain 'This is a test'

---

In this example, we echo the text, 'This is a test' to fluent's stdin input. This is combined with the input question and sent to the OpusChain. The two will be combined if I include a string in the request.

echo "This is a test" | fluent OpusChain ''

---

In this request, we combine 'What is this?' with 'This is a test' and send it to the LLM.

echo 'this is a test' | fluent OpusChain 'What is this? '

Any text can be piped into fluent through stdin. However, token size limits do apply. If you exceed the size limit, you will receive a response like this:

{"statusCode":500,"success":false,"message":"Error: predictionsServices.buildChatflow - 400 {\"type\":\"error\",\"error\":{\"type\":\"invalid_request_error\",\"message\":\"prompt is too long: 202122 tokens > 199999 maximum\"}}","stack":{}}

---

Adding file contents to the request payload

fluent OpusChain 'analyze these data points. ' --additional-context-file /Users/n/Downloads/Financial Sample.csv