Claude Skill · Visibility & GEO

AI Visibility Audit

Traditional SEO asks do we rank. This asks are we cited, and are we described correctly — across ChatGPT, Claude, Perplexity, Gemini, and Google AI Overviews. Different mechanics, different fixes.

Language · Markdown License · MIT Type · Claude Skill

View on GitHub north_east Fork it fork_right

01 · The Problem

Why This Exists

Rank tracking cannot see the surface that is eating your clicks. A brand can hold position three for its head term and still be absent from every AI answer that term now triggers.

The failure is almost never content volume. It is that the sources those answers cite do not mention you — so publishing more owned content moves nothing.

This skill measures presence, diagnoses which of five root causes is responsible, and returns a ranked fix list you can hand to someone on Monday.

02 · Method

How It Works

01

Build the prompt set

20–30 prompts across five intent tiers — discovery, comparison, problem-first, branded, and buying-criteria — weighted toward the unbranded questions where visibility is actually won.

02

Run & record

For each prompt: is the brand mentioned, where, described how — and crucially, which sources the answer cites. That citation column is where the fix lives.

03

Diagnose the root cause

Every gap maps to one of five causes: citation absence, entity ambiguity, extractability, comparative absence, or factual drift. Naming the cause is what makes the fix specific.

04

Ship a ranked fix list

Prioritized by (impact × confidence) ÷ effort, each item states what to do, why it moves AI visibility, who owns it, and how you'll know it worked. Re-measure in 60 days.

03 · Run It

Two Minutes To First Output

# 1. Clone the repo
git clone https://github.com/seosdm/ai-visibility-audit.git

# 2. Copy SKILL.md into a Claude Project, then prompt:
Run the ai-visibility-audit skill for [brand].
Category: [what they sell]. Competitors: [3–5].

# 3. Re-run the same prompt set every 60 days to measure movement.

No API keys, no setup. The skill is a SKILL.md file plus a worked example — it runs anywhere Claude does.

04 · Proof

Sample Output

Presence Rate by Intent Tier Example · anonymized

78% Branded

31% Comparison

12% Problem-first

6% Discovery

Verdict: strong when named, near-invisible in discovery. Root cause — the three listicles cited across discovery prompts don't mention the brand. Highest-leverage move is earning placement in those sources, not publishing more owned content.

See the full worked example in /examples.

If The Tool Gets You 60% There

I'll Teach You The Rest

Free skill, real judgment behind it. Want it run on your business, or your team taught to run it themselves?

Work With Me On Visibility north_east