FREE CLAUDE SKILL · METADATA + SERP CTR
A free Claude skill that pulls the live top 10 for your target keyword, reads what every competitor claims, and writes titles and descriptions that claim what none of them do. Then exports the whole thing as a spreadsheet sorted by what to fix first. Built by Lil Big Things, the GEO agency for B2B tech.
Runs in Claude · Connects to Ahrefs for live SERP and keyword data · Exports CSV and XLSX · No card required
The problem
Every metadata tool tells you the same three things: your title is too long, your description is too short, your keyword is missing. All true, all worth fixing, and none of it explains why a buyer scanning ten results would click yours.
The thing that decides the click is what the other nine say. If six of them use the same phrase you do, matching that phrase makes you invisible. The opportunity is whatever nobody claims.
Real SERP data, 22 July 2026
SERP: "webflow development services" #2 Webflow Development Services #3 Webflow Development Services #4 Webflow Development Service #5 Webflow Web Development Services
GAP: No segment. No timeline. No proof. No price.
The only differentiator in the entire top 10
is one competitor adding a city name.
Six of ten titles were the bare phrase. A title naming the buyer segment and a proof number has that SERP to itself.
What it does
Pulls the live top 10
Reads every competitor title for your target keyword and identifies what they all say and what none of them say.
Maps commercial intent
Classifies every page as Money, Capture, Proof, Discovery, or Utility. Metadata effort follows revenue, not traffic.
Scores lead value
Ranks by CPC, difficulty against your actual domain authority, and distance to conversion. A KD 1 term at $5 a click beats a KD 57 term at ten times the volume.
Catches the intent traps
Some high-volume keywords serve people who want to build the thing, not buy it. The skill reads the SERP page types and flags those AVOID rather than sending you after traffic that never converts.
Finds the technical defects
Missing and misdirected canonicals, Open Graph tags copy-pasted between pages, duplicate titles. These are breakage, not optimisation, and they get reported separately.
Specifies internal links
From-page, to-page, exact anchor text. Replacing "Learn more" with a keyword-bearing anchor is usually the cheapest ranking improvement on a site.
The output
Not a report you read once. A working document your team can move through row by row, in Google Sheets or Excel.
Terminal demo
What makes it different
Reads the SERP, not a rulebook
Every proposed title is checked against the live top 10. If a competitor already owns the angle, the skill picks a different one.
Sorts by revenue, not traffic
CPC, difficulty against your real authority, and distance to conversion. The person implementing works top-down and may stop halfway, so the money goes at the top.
Flags the traps
High-volume keywords that serve DIY intent get marked AVOID with the reason, instead of quietly padding the opportunity list.
Separates defects from opinions
A canonical pointing at a dead URL is breakage. A four-character title overrun is a preference. They are reported differently because they need different urgency.
Never invents proof
A number in a title is powerful because a buyer can check it. The skill uses figures you actually have and marks the gaps rather than filling them.
Exports to a real spreadsheet
Six tabs, frozen headers, autofilter, colour-coded flags. Opens cleanly in Google Sheets and Excel. Hand it to whoever does the implementing.
How it works
Install the skill
Download and save into Claude. It triggers whenever you mention metadata, title tags, meta descriptions, CTR, or SERP appearance.
Point it at your pages
Give it a domain, a sitemap, or a URL list. It classifies by tier, fetches live metadata, and pulls the SERP for every commercial term.
Get the spreadsheet
CSV and XLSX. Sorted by priority, with the reasoning for every recommendation in the row.
Who it's for
Want us to run it and implement the changes? That is the Lil Big Things GEO programme. Book a GEO strategy call
What's in the file
meta-optimizer.skill The installable skill. Save once, run on demand.
meta-optimizer/
SKILL.md SERP differential method, intent mapping,
lead scoring, technical checks, output spec.
references/
serp-rules.md Length limits, truncation behaviour, OG
specifics, commercial vs editorial metadata.Honest note
This grades and rewrites metadata. It will not fix a page that has nothing to say. If a title cannot claim something the top 10 does not, the problem is the offer, not the tag. The skill will tell you that rather than writing a cleverer sentence around it.
More free skills
SEO + AEO Audit
Checks whether engines can crawl, render, and cite your site.
Homepage Build + Audit
Grades your homepage against six buyer questions and rewrites what is broken.
AEO Blog Post
Writes content engineered to get cited.
Content Brief
Builds a brief a writer can execute without a follow-up call.
Meta Description Optimizer (this)
Rewrites title tags and descriptions against the live top 10.
GET THE SKILL
Drop your email and we will send the skill file plus a short guide to running your first SERP differential. We will also send the occasional GEO Times update. Unsubscribe anytime.
FAQ