Skip to main content
Analyze your brand’s writing style from existing blog posts and create a detailed voice profile.

​
Overview

extract_tone builds a brand voice profile:
  • Automated analysis — reads up to 10 blog posts to understand your style
  • Voice characteristics — identifies tone, vocabulary, sentence structure, and patterns
  • Reusable profile — stores profile for use in rewrite_article
  • Comprehensive report — detailed breakdown of your writing style
This tool requires an API key and costs 500 credits per extraction.

​
Parameters

​
blog_url
string
required
URL of your blog or blog listing page. The tool will crawl to find individual posts.
​
customer_id
string
Customer identifier for profile storage. If omitted, uses your API key.
​
max_articles
number
default:"10"
Maximum number of articles to analyze (1-10). More articles = more accurate profile.

​
Example prompts

Extract the brand voice from https://example.com/blog

What's my writing style?

Analyze my brand tone from my blog posts

​
Response structure

The tool returns a JSON object with:
​
tone_profile
object
Complete voice profile
​
articles_analyzed
number
Number of articles actually analyzed
​
profile_id
string
Unique identifier for this tone profile (for use in rewrite_article)
​
created_at
string
When the profile was created

​
Usage example

# How the tool is called in the MCP server
@mcp.tool(annotations=WRITE)
def extract_tone(
    blog_url: str,
    customer_id: str | None = None,
    max_articles: int = 10,
) -> str:
    """Analyze your brand's writing style from blog posts."""
    client = _get_client()
    return _call(
        client.extract_tone, blog_url, customer_id=customer_id, max_articles=max_articles
    )

​
What gets analyzed

The tool examines:
  • Tone and voice — formal vs casual, serious vs playful, authoritative vs conversational
  • Vocabulary — word choice, jargon usage, reading level
  • Sentence structure — length, complexity, variety
  • Paragraph patterns — typical paragraph length and flow
  • Content structure — how you organize information
  • Stylistic choices — use of questions, examples, metaphors, humor
  • Formatting habits — heading hierarchy, list types, emphasis

​
When to use

Use extract_tone when:
  • Before using rewrite_article — establish your voice first
  • New content team members — codify your brand voice for consistency
  • Rebranding — document your target voice
  • Content audits — understand current voice across all content
Run this once and reuse the profile for all article rewrites. You only pay 500 credits once, then use the profile unlimited times.

​
Analysis quality

Better profiles come from:
  • More articles — 10 articles gives most accurate results
  • Recent content — reflects current brand voice
  • Consistent authors — single voice is easier to profile
  • Long-form content — blog posts with 1000+ words
Avoid analyzing:
  • Press releases (too formal)
  • Product descriptions (too promotional)
  • Mixed author content (inconsistent voice)
  • Very short posts (not enough signal)

​
Profile storage

Your tone profile is stored and reusable:
  • Profiles are tied to your API key (or custom customer_id)
  • Multiple profiles supported (for different brands/products)
  • Profiles don’t expire
  • Update by running extract_tone again (creates new profile)

​
Using the profile

After extraction, use the profile in rewrite_article: 
Rewrite this article using my brand voice
The rewrite tool automatically finds your most recent tone profile and applies it.

​
Cost

Extracting a tone profile costs 500 credits. This includes:
  • Crawling and reading up to 10 blog posts
  • Natural language analysis of writing patterns
  • Profile generation and storage
  • Unlimited future use in rewrites

​
Next steps