graph TD %%{init: {'theme': 'mc','layout': 'elk'}}%% Prompt-uq0mq[<div><img alt="logo" src="/_astro/square-terminal.BMOXc-nZ.svg" style="height: 20px !important;width: 20px !important"/></div>Prompt Propuesta de Imagen] style Prompt-uq0mq stroke:#a170ff Prompt-cf2w7[<div><img alt="logo" src="/_astro/square-terminal.BMOXc-nZ.svg" style="height: 20px !important;width: 20px !important"/></div>Prompt Resumen del artículo] style Prompt-cf2w7 stroke:#a170ff Prompt-8b4xy[<div><img alt="logo" src="/_astro/square-terminal.BMOXc-nZ.svg" style="height: 20px !important;width: 20px !important"/></div>Prompt Redactor de tweet] style Prompt-8b4xy stroke:#a170ff OpenAIModel-k3h4h[<div><img alt="logo" src="/_astro/openAI.BhmuxEs3.svg" style="height: 20px !important;width: 20px !important"/></div>OpenAI] style OpenAIModel-k3h4h stroke:#a170ff URL-svy97[<div><img alt="logo" src="/_astro/layout-template.DBcxlmmt.svg" style="height: 20px !important;width: 20px !important"/></div>URL] style URL-svy97 stroke:#a170ff ChatOutput-5l8gy[<div><img alt="logo" src="/_astro/messages-square.BaSDmT6g.svg" style="height: 20px !important;width: 20px !important"/></div>Chat Output] style ChatOutput-5l8gy stroke:#a170ff DallEImageGenerator-jhlv0[<div><img alt="logo" src="/_astro/openAI.BhmuxEs3.svg" style="height: 20px !important;width: 20px !important"/></div>DALL·E Image Generator] style DallEImageGenerator-jhlv0 stroke:#a170ff OpenAIModel-nx97c[<div><img alt="logo" src="/_astro/openAI.BhmuxEs3.svg" style="height: 20px !important;width: 20px !important"/></div>OpenAIPropuesta de Imagen] style OpenAIModel-nx97c stroke:#a170ff Prompt-hse57[<div><img alt="logo" src="/_astro/square-terminal.BMOXc-nZ.svg" style="height: 20px !important;width: 20px !important"/></div>Prompt Generar Imagen] style Prompt-hse57 stroke:#a170ff JsCode-bst4b[JS Code] style JsCode-bst4b stroke:#a170ff CombineText-46th0[Combine Text] style CombineText-46th0 stroke:#a170ff ParseData-tpkkd[<div><img alt="logo" src="/_astro/braces.Djq0PW4_.svg" style="height: 20px !important;width: 20px !important"/></div>Parse Data2] style ParseData-tpkkd stroke:#a170ff ParseData-4oihd[<div><img alt="logo" src="/_astro/braces.Djq0PW4_.svg" style="height: 20px !important;width: 20px !important"/></div>Parse Data3] style ParseData-4oihd stroke:#a170ff OpenAIModel-52v7e[<div><img alt="logo" src="/_astro/openAI.BhmuxEs3.svg" style="height: 20px !important;width: 20px !important"/></div>OpenAIPropuesta de Imagen] style OpenAIModel-52v7e stroke:#a170ff Prompt-8b4xy -.- OpenAIModel-k3h4h linkStyle 0 stroke:#a170ff Prompt-uq0mq -.- OpenAIModel-nx97c linkStyle 1 stroke:#a170ff OpenAIModel-nx97c -.- Prompt-hse57 linkStyle 2 stroke:#a170ff Prompt-hse57 -.- DallEImageGenerator-jhlv0 linkStyle 3 stroke:#a170ff DallEImageGenerator-jhlv0 -.- JsCode-bst4b linkStyle 4 stroke:#a170ff JsCode-bst4b -.- ParseData-tpkkd linkStyle 5 stroke:#a170ff ParseData-tpkkd -.- CombineText-46th0 linkStyle 6 stroke:#a170ff CombineText-46th0 -.- ChatOutput-5l8gy linkStyle 7 stroke:#a170ff OpenAIModel-k3h4h -.- CombineText-46th0 linkStyle 8 stroke:#a170ff URL-svy97 -.- ParseData-4oihd linkStyle 9 stroke:#a170ff ParseData-4oihd -.- Prompt-cf2w7 linkStyle 10 stroke:#a170ff Prompt-cf2w7 -.- OpenAIModel-52v7e linkStyle 11 stroke:#a170ff OpenAIModel-52v7e -.- Prompt-8b4xy linkStyle 12 stroke:#a170ff OpenAIModel-52v7e -.- Prompt-uq0mq linkStyle 13 stroke:#a170ff

Documentation – Generate content from a URL

🧩 Overview

The workflow automatically transforms the text of a web page into a concise article summary and a complementary image. It then produces a short, engaging tweet that incorporates the summary and the image, delivering both the textual content and the image in a single chat‑style output. This end‑to‑end pipeline saves time by replacing manual summarization, image creation, and tweet drafting with AI‑powered automation.

⚙️ Main Features

Automatic web‑page extraction – fetches raw content from any supplied URL.
Text parsing – converts the fetched data into clean, plain text.
Dynamic summarization – generates a brief, key‑point summary of the article.
Tweet composition – creates a concise, emoji‑rich tweet based on the summary.
Image generation – produces a relevant illustration using a generative image model.
Markdown conversion – turns the binary image into Markdown syntax for display.
Content merging – concatenates the tweet and image description into a single message.
Chat output – presents the final text and image in a chat‑friendly format.

🔄 Workflow Steps

Component Name	Role in the Workflow	Key Inputs	Key Outputs
URL	Fetches web page content	URL (string)	Data (raw page content)
Parse Data	Extracts plain text from fetched data	Data (raw page content)	Text (article body)
Prompt (Resumen del artículo)	Builds a prompt to ask for a summary	Article text	Prompt message
OpenAI (Summarizer)	Generates a concise summary of the article	Prompt message	Text (article summary)
Prompt (Redactor de tweet)	Builds a tweet‑generation prompt using the summary	Article summary	Prompt message
OpenAI (Tweet)	Generates an engaging tweet with emojis	Prompt message	Text (tweet)
Prompt (Propuesta de Imagen)	Builds an image‑generation prompt using the summary	Article summary	Prompt message
DALL·E Image Generator	Creates an image that illustrates the article	Prompt message	Binary (image file)
JS Code	Converts the binary image into a Markdown‑formatted string	Binary image	Data (Markdown image link)
Parse Data	Extracts the Markdown string into plain text	Data (Markdown link)	Text (image description)
Combine Text	Concatenates the tweet and the image description	Tweet text, Image description	Text (combined message)
Chat Output	Displays the final combined message in a chat view	Combined text	Chat message (visible output)

🧠 Notes

Credential Management – The workflow relies on valid OpenAI and DALL·E API keys; these must be securely stored and referenced by the relevant model components.
Prompt Length Constraints – Image‑generation prompts are limited to 4000 characters; ensure the summary stays within this limit.
Rate‑Limiting – The image generator and text model each have request quotas; monitor usage to avoid exceeding limits.
Data Privacy – The fetched page content is treated as transient data and is not stored beyond the scope of the workflow unless explicitly configured.
Error Handling – If any component fails (e.g., network timeout, API error), the workflow will halt and report the failure; graceful degradation is not implemented by default.
Output Formatting – The JS Code component assumes the image is returned as a base‑64 encoded binary; if the image format changes, the conversion logic may need adjustment.
Scalability – While the current design processes a single URL at a time, the structure can be extended to batch URLs by adding a mapping mode to the Parse Data or Image Generator components.