Go to Studio

Tag Extractor

Easily extract specific content from HTML or XML tags in your text and AI outputs.

What does this node do?

The Tag Extractor node analyzes an input text and extracts the content inside specific HTML or XML tags (wrapped in angle brackets < and >).

Common use cases:

  • Pull out specific sections generated by an LLM (e.g. <score_seo>85/100</score_seo>).
  • Extract standard elements from a previously scraped web page (e.g. h1, h2, p).
  • Structure complex outputs by isolating precise variables to pass on to other tools in your workflow.
Tag Extractor node on the Draft & Goal Studio canvas

Quick setup

You’ll find the Tag Extractor node in the left panel under Utilities > Text Processing.

Add the node to your workflow

Drag and drop the Tag Extractor node onto your workspace.

Connect the input

Connect the output of an upstream node (e.g. an LLM, AI Agent, or Web Scraper) to the Tag Extractor input.

Open settings

Click the node’s gear icon (Tool Settings) to open the configuration panel.

Set the target tag

In the Tag field, enter the exact tag name to extract without angle brackets (e.g. type h2, not <h2>).

Choose the output format

Select plain text (Text) or structured (Array) depending on how you’ll use the result in the rest of the workflow.

Configuration

Configure the node by specifying the target tag and how extraction should behave.

Example Tag Extractor configuration (tag, format, error handling)

Required parameters

Tag string required

Tag — The exact HTML or XML tag name to target, without angle brackets. For example, to extract <content_html>, enter content_html.

Response Format string required default: Text

Response Format — How the extracted data is returned:

  • Text: Plain text. If multiple identical tags are found, their contents are separated by line breaks.
  • Array: Returns a JSON list (0, 1, 2…). Use this when you need to iterate over results with a Loop node.
Error Handling string required default: None

Error Handling — What happens when the tag is not found:

  • None: The workflow stops and throws an error (default).
  • Skip and continue: The node ignores the error, returns an empty value, and the workflow continues.

What does the node output?

The Tag Extractor returns the content inside the specified tag. The exact format depends on the Response Format setting.

Using the output in other nodes

To use this node’s result in another node (e.g. a merge or LLM), use the variable syntax with double curly braces, for example: {{tag_extractor_h2}}.

Tag Extractor output in use (variable used in another node)

Example: automated SEO article writing

The Tag Extractor is especially useful when combined with an LLM to structure its output in a deterministic way.

  1. The prompt (LLM node) — Ask the AI to write an article and wrap its response in exact XML tags. 
    Write an SEO article. You MUST wrap your response with these exact tags:
<keyword>The main keyword here</keyword>
<html-content>The article HTML here</html-content>
<score-copywriting>Your self-assessment score out of 100 here</score-copywriting>
  2. Extraction (Tag Extractor nodes) — Add 3 Tag Extractor nodes after the LLM:
    • The 1st extracts the keyword tag and sends it to SEMrush.
    • The 2nd extracts the html-content tag and sends it to the WordPress node.
    • The 3rd extracts the score-copywriting tag to store in a Google Sheet.

How it fits into a workflow

Here are two common workflow patterns using the Tag Extractor.

flowchart LR
    A[LLM outputs text + tags] --> B[Tag Extractor: keyword]
    A --> C[Tag Extractor: content_html]
    B --> D[SEO / Analytics tools]
    C --> E[WordPress / CMS]

Pattern 2: Web page extraction

flowchart LR
    A[Web Scraper] --> B[HTML to Markdown]
    B --> C[Tag Extractor: h2]
    C --> D[Loop: Process each H2]

Best practices and pitfalls

  • Strict prompts: In your AI prompts, insist that the model use the exact tag name and never change it.
  • Use Array for lists: When extracting repeated tags like h2 or li, prefer the Array format so you can process each item cleanly.
  • Don’t include angle brackets: In the node config, enter summary not <summary>.
  • Don’t target tags with attributes: The node doesn’t support CSS classes or IDs. A tag like <h2 class="title"> won’t be parsed correctly by the Tag Extractor.
Tip

To make automated workflows more resilient, set Error Handling to Skip and continue when the tag is optional, so the process doesn’t fail entirely.

Warning

Tags with attributes are not supported: Complex HTML tags with styles or classes (e.g. <div style="color:red;">) are not handled. Stick to plain tags (standard or custom).

Common issues and fixes

The node finds no tag and the workflow stops

Cause: The AI (LLM) didn’t output the tag exactly as requested, or added extra spaces or different casing.

Fix: In your LLM node, tighten the instructions (e.g. “You must use the exact format <score> with no extra characters or changes in capitalization.”).

I can't extract an HTML tag from a scraped web page

Cause: The page uses HTML tags with attributes (e.g. <h1 id="main-title" class="header">). The Tag Extractor only reads simple tags.

Fix: Use an HTML Cleaner or HTML to Markdown node first to strip attributes before sending the content to the Tag Extractor.

All results appear on one line

Cause: There are multiple identical tags in the text and Response Format is set to Text with no clear separator.

Fix: Set Response Format to Array. You’ll get a clean list that’s easy to work with.

See how the Tag Extractor works with other nodes:

LLM

Generate text wrapped in specific XML tags for later extraction.
Web Scraper

Fetch raw page HTML to extract specific elements.
WordPress Post Create

Send the content of the <html-content> tag straight to your site.
Loop

Iterate over each result when using the “Array” output format.