Skip to main content

Blog writer

Build a Blog Writer flow for a one-shot application using OpenAI.

This flow extends the Basic Prompting flow with the URL and Parser components that fetch content from multiple URLs and convert the loaded data into plain text.

OpenAI uses this loaded data to generate a blog post, as instructed by the Text Input and Prompt components.

Prerequisites​

Create the blog writer flow​

  1. From the Langflow dashboard, click New Flow.
  2. Select Blog Writer.
  3. The Blog Writer flow is created.

This flow creates a blog article generator with Prompt, OpenAI, and Chat Output components, augmented with reference content and instructions from the URL and Text Input components.

The URL component extracts raw text and metadata from one or more web links. The Parser component converts the data coming from the URL component into plain text to feed the prompt.

To examine the flow's prompt, click the Template field of the Prompt component.


_10
Reference 1:
_10
_10
{references}
_10
_10
---
_10
_10
{instructions}
_10
_10
Blog:

The {instructions} value is received from the Text Input component, and one or more {references} are received from a list of URLs parsed from the URL component.

Run the blog writer flow​

  1. Click the Playground button, and then click Run Flow. A blog post about Langflow is generated, with content sourced from langflow.org and docs.langflow.org.
  2. To write about something different, change the values in the URL component and adjust the instructions on the left side bar of the Playground. Try again and see what the LLM constructs.
Search