Fifteen diagnostic skills for the tracking layer everything else depends on: GTM, GA4, Pixel and CAPI, consent, click IDs and duplicate events.
.Bad tracking is worse than no tracking.
No tracking forces humility. Bad tracking lets you make confident, expensive mistakes.
Join the waitlist and get notified when the new AdLume is ready. It catches wasted ad spend automatically and optimises campaigns around the clock.
One email, on the day access opens. Nothing else, and you can leave the list with one click.
You scale budget because ROAS looks healthy, but the real source of revenue is hidden. You cut a channel because it looks broken, but the pixel is double-firing. You report a 40% lift when the event fired twice.
That is why this pack exists: 15 Claude Skills for GTM, GA4, Google Ads, Meta Pixel + CAPI, consent, click IDs and offline conversions.
Use this when tracking quality affects budget, reporting or platform decisions.
Free library for performance marketers
Templates for Claude, GTM, GA4, Google Ads, Meta Pixel + CAPI and consent mode. Built for the boring but expensive part: checking whether the signal is actually clean.
Free. Open source. Use one skill or fork the whole pack.
The skills are grouped by the kind of tracking problem you are trying to solve.
A Claude Skill is a self-contained workflow. It tells Claude when to activate, what evidence to request, how to analyze it, and what output to return.
That matters because tracking audits are not normal brainstorming tasks.
The model needs to know what event should fire, where it should fire, what can break it, and when missing evidence changes the conclusion.
The first version of this pack is diagnostic-first. It does not need write access to your containers, and it does not pretend AI should silently fix tracking for you.
Start with the data you actually have. Screenshots are fine for a first pass, but exports give the highest-confidence output.
If the data is missing, the skill should say so. That is the point. Tracking work gets risky when AI sounds confident while the signal is dirty.
Before you change a tag
If the account looks wrong, do not start with optimization. Start with the signal path: event, source, deduplication, consent and reporting confidence.
Open the tracking audit skills ->
Useful when ROAS looks good but the numbers do not feel real.
The full library of AdLume skills: paid ads, landing pages, B2B leads, GTM tracking, e-commerce.
The bundle goes straight to your inbox, then one email a week. Unsubscribe in one click. We never share your address.
Use this when Meta, Google Ads, GA4 or your CRM disagree on what happened.
The Duplicate Event Finder looks at:
The output is not a vibe check. It should return a structured view:
Event | Source | Fires | Deduplicated? | Risk | Fix | Confidence |
|---|
That turns “something is wrong with tracking” into a specific fix.
Use this after a cookie banner update, CMP migration or unexplained conversion drop.
The Consent Mode Impact Review estimates:
It does not recommend ignoring consent. It helps you understand what the signal loss means for your decisions.
The skills do not.
They are built for diagnosis and approval-ready recommendations.
Weekly operating rhythm
The goal is not a one-off audit. The goal is a repeatable signal quality habit before scaling spend or trusting platform recommendations.
Run it read-only first. Anything that changes tracking still needs approval.
Clone the repo.
git clone https://github.com/mardab96/gtm-tracking-claude-skills.git
Copy the skill folders into your user-level skills directory:
cp -r gtm-tracking-claude-skills/* ~/.claude/skills/
Start a new Claude Code session. The skills activate automatically when their description matches what you are asking for.
Each skill is plain Markdown with YAML frontmatter. Paste the relevant SKILL.md file into context when you want to use it.
Three practical starting points:
If you also run broader paid ads, you may want the companion packs:
The broader shift is simple: tracking is becoming a set of repeatable AI-assisted diagnostics.
The teams that win will keep the messy human parts visible: business context, signal quality, tradeoffs, approvals and the reason behind each change.
If your ROAS, CPA or conversion volume looks suspicious, start with the signal. Bad tracking makes every other optimization look smarter than it is.
Free, open source, ready to adapt
Bad tracking makes every optimization look smarter than it is. Start with one messy part of the setup and turn the fix into a repeatable check.
Download the GTM tracking skills ->
No form. No gate. Just the repo.
AdLume watches your ad accounts and brings the next fix to your team's tools.
One email, on the day access opens. Nothing else, and you can leave the list with one click.
No. The pack is diagnostic-first: the skills do not edit GTM containers, publish tags, change conversion actions, modify consent settings or touch data layer code. They produce diagnosis and approval-ready recommendations.
What data do I need to run a tracking audit skill?Start with what you have: a GTM container export or screenshots, GA4 event or key event reports, and Google Ads conversion action settings. Screenshots are fine for a first pass; exports give the highest-fidelity read. If data is missing, the skill says so instead of guessing.
How do I install the skills in Claude Code?Clone the GitHub repo, copy the skill folders into your user-level skills directory (~/.claude/skills/), and start a new Claude Code session. The skills activate automatically when their description matches what you ask for.
Is the skill pack free?Yes. It is free and open source, with no form and no gate. Use one skill, adapt the required-data lists to your stack, or fork the whole pack.