For Agents · MCP

Your task.
The right skill.

The tunnel uses the same cards as the catalogue. Browse only as deep as needed — or load a broad bundle immediately.

No local installationNo M11 loginRead-only

Broad task

Load 5–10 and go

SEO, Sales, Agents or another broad area → one bundle call → work.

MCP endpoint: https://skills.m11.ch/mcp

Read-only access to published skills. Default 8, maximum 10 skills / 120,000 characters.

Task Packs

Marketing Foundation

Compact two-skill starter: clarify positioning and choose a lead magnet. Use Marketing Launch for the broader eight-skill go-to-market workflow.

Marketing Launch

Build an evidence-led marketing plan from ICP and competition through positioning, campaigns, growth and measurement.

Agent Audit Essentials

Diagnose architecture and context, plan agent-team responsibilities, then organize project context and session handoffs. Memory and cost-runtime reviews remain outside this pack.

Conversion and Activation Review

Review the journey from landing page and lead capture through registration, first value and transparent upgrades.

Campaign Content and Brand Review

Plan a campaign, draft its channel content and review the work against actual brand guidance.

Content and Launch Planning

Prioritize an editorial roadmap and plan how to launch and distribute it across suitable channels.

Customer, Market and Community Research

Understand customer needs, compare competitors and plan a community around real member value.

Lead Magnet to Lifecycle Nurture

Choose a relevant lead magnet, then draft a permission-based nurture journey with entry, suppression and exit rules.

API Contract and Observability

Define the API contract, then plan how to observe its latency, failures and retries. Guidance and checklist; no production changes.

Data Analysis Foundation

Profile a dataset, choose and interpret statistical methods, then validate calculations and conclusions before sharing.

Choose an area

01 · Category

What the MCP returns at this step

35 · Marketing · Paid Advertising

Cross-Platform Advertising Review

Original multi-platform ad-audit orchestration; required platform packages and scoring files are not included.

adsreviewcross-platformwerbekanalvergleichwerbeplattformen

Cross-Platform Marketing Audit · Original SKILL.md

---
name: cross-platform-audit
description: "Unified multi-platform paid media audit with parallel execution. Spawns scored audits for Google Ads, Meta Ads, and Microsoft Ads simultaneously, merges results into a budget-weighted aggregate health score (A-F), adds cross-platform analysis (budget allocation, tracking consistency, creative alignment, attribution overlap), and produces a prioritized action plan. Optionally generates a client-ready deck via pro-deck-builder. Use when the user asks for a full paid media audit, cross-platform review, or multi-channel ad assessment."
license: MIT
origin: custom
author: Rebecca Rae Barton
author_url: https://github.com/thatrebeccarae
metadata:
version: 1.0.0
category: paid-media
domain: cross-platform
updated: 2026-03-16
tested: 2026-03-17
tested_with: "Claude Code v2.1"
---

# Cross-Platform Audit — Unified Paid Media Assessment

Orchestrates parallel scored audits across Google Ads, Meta Ads, and Microsoft Ads, then merges results into a single budget-weighted health score with cross-platform intelligence.

## Install

```bash
git clone https://github.com/thatrebeccarae/claude-marketing.git && cp -r claude-marketing/skills/cross-platform-audit ~/.claude/skills/
```

## Context Intake (Always Do First)

Before running audits, collect this information. Without it, benchmarks will be generic and recommendations may be wrong.

Ask these questions (combine into one prompt):

1. **Business type** — E-commerce/DTC, SaaS/B2B, Local Service, Lead Gen, Agency, Other
2. **Monthly ad spend** — Total and per-platform breakdown (approximate is fine)
3. **Primary goal** — Revenue/ROAS, Leads/CPA, App Installs, Brand Awareness
4. **Active platforms** — Which platforms are they advertising on? (Google, Meta, Microsoft, others)
5. **Data available** — What can they provide? (exports, screenshots, pasted metrics, MCP access)

Use the provided context to:
- Determine which platform audits to run (skip platforms they don't use)
- Calculate budget share per platform for aggregate scoring
- Calibrate severity based on spend level ($5K/mo gets different advice than $100K/mo)

## Orchestration Workflow

### Step 1: Collect Context
Gather business type, platforms, budgets, goals, and available data per the intake above.

### Step 2: Spawn Parallel Audits

Use the Task tool to run platform audits simultaneously. Each audit task should:

1. Load `skills/shared/scoring-system.md` for the scoring algorithm
2. Load the platform-specific `CHECKS.md` for the audit checklist
3. Load the platform-specific `SKILL.md` and `REFERENCE.md` for diagnostic context
4. Evaluate each applicable check as PASS, WARNING, or FAIL
5. Calculate the platform health score using weighted formula
6. Identify Quick Wins
7. Write results to a structured output

**Spawn these tasks in parallel** (adjust based on which platforms the user has):

```
Task 1: Google Ads Audit
→ Read skills/shared/scoring-system.md
→ Read skills/google-ads/CHECKS.md
→ Evaluate 74 checks against provided data
→ Calculate score → Write google-audit-results.md

Task 2: Meta Ads Audit
→ Read skills/shared/scoring-system.md
→ Read skills/facebook-ads/CHECKS.md
→ Evaluate 46 checks against provided data
→ Calculate score → Write meta-audit-results.md

Task 3: Microsoft Ads Audit
→ Read skills/shared/scoring-system.md
→ Read skills/microsoft-ads/CHECKS.md
→ Evaluate 30 checks against provided data
→ Calculate score → Write microsoft-audit-results.md
```

If the user only has 2 platforms, spawn only those 2 tasks.

### Step 3: Merge Results

After all audit tasks complete:

1. Read all `*-audit-results.md` files
2. Extract per-platform scores and grades
3. Calculate budget-weighted aggregate score:
```
Aggregate = Σ(Platform_Score × Platform_Budget_Share)
```
4. Run cross-platform analysis (see Cross-Platform Analysis below)
5. Merge Quick Wins from all platforms into a unified priority list
6. Generate the unified report

### Step 4: Cross-Platform Analysis

Add intelligence that no single-platform audit can provide:

**Budget Allocation Analysis**
- Is spend distributed optimally across platforms?
- Are high-performing platforms budget-constrained while low performers have excess?
- Does the platform mix match the business type? (e.g., B2B without LinkedIn, ecom without Shopping)
- Compare platform CPAs and flag reallocation opportunities

**Tracking Consistency**
- Are conversion events defined consistently across platforms?
- Is revenue/value tracking active on all platforms or only some?
- Are attribution windows aligned? (7-day click on Meta but 30-day on Google = inconsistent reporting)
- Is there a single source of truth (GA4) or are platforms reporting independently?

**Creative Alignment**
- Is messaging consistent across platforms or contradictory?
- Are creative formats appropriate per platform? (vertical video for Meta/TikTok, text-heavy for Search)
- Is creative being refreshed at similar cadences or is one platform stale?

**Attribution Overlap**
- Are platforms double-counting conversions? (same user converts, Google and Meta both claim credit)
- What percentage of total reported conversions exceeds actual conversions?
- Is there a de-duplication strategy (UTMs + GA4 as arbiter)?

**Kill List**
- Campaigns/ad sets across ALL platforms with CPA >3x target
- Aggregate wasted spend across platforms
- Quick calculation: total monthly waste = Σ(flagged campaign spend)

**Scaling Opportunities**
- Which campaigns across all platforms have CPA below target AND are budget-constrained?
- What is the estimated incremental revenue from reallocating waste to winners?

### Step 5: Generate Unified Report

Output the unified report in the format specified in EXAMPLES.md.

### Step 6: Deck Generation (Optional)

If the user requests a deck (`--deck` flag, or asks for a presentation/deliverable):
- Take the unified report output
- Pipe to pro-deck-builder skill
- Generate a branded HTML slide deck with:
- Executive summary slide (aggregate score, grade, key findings)
- Per-platform score cards
- Cross-platform analysis slides
- Quick Wins slide
- Prioritized action plan slides
- Budget reallocation recommendation slide

## Hard Rules

These constraints apply across all platforms and must never be violated:

1. **Never recommend changes during active learning phases** on any platform.
2. **3x Kill Rule is universal** — flag ANY campaign on ANY platform with CPA >3x target.
3. **Tracking must be verified before optimization recommendations** — if conversion tracking is broken, that is the only recommendation until fixed.
4. **Budget reallocation recommendations must account for minimum thresholds** — don't recommend shifting $500/mo to a platform that needs $3K/mo minimum.
5. **Attribution overlap must be flagged** — if total platform-reported conversions exceed GA4 by >20%, call it out.
6. **Special Ad Categories apply cross-platform** — if housing/employment/credit/finance on one platform, compliance checks apply to all platforms.

## Relationship to Other Skills

This skill **orchestrates** the following skills — it does not replace them:

| Skill | Role in Orchestration |
|-------|----------------------|
| `google-ads` | Provides CHECKS.md + diagnostic frameworks for Google audit task |
| `facebook-ads` | Provides CHECKS.md + diagnostic frameworks for Meta audit task |
| `microsoft-ads` | Provides CHECKS.md + diagnostic frameworks for Microsoft audit task |
| `account-structure-review` | Complementary — run separately for deep structural analysis |
| `wasted-spend-finder` | Complementary — run separately for detailed waste analysis with CSV output |
| `shared/scoring-system.md` | Provides scoring algorithm used by all audit tasks |
| `pro-deck-builder` | Optional downstream — generates client-ready presentation |

## How to Use This Skill

Ask questions like:
- "Run a full audit across all my ad platforms"
- "Score my Google and Meta accounts and give me one unified health grade"
- "Audit my paid media and build a deck I can present to my client"
- "I spend $20K on Google, $15K on Meta, and $5K on Microsoft — audit everything"
- "Which platform should I shift budget to based on performance?"
- "Give me a cross-platform audit with a prioritized action plan"

When to use

Original multi-platform ad-audit orchestration; required platform packages and scoring files are not included.

What you get

The structured plan or review specified in the original, with evidence and open questions.

How it works

Confirm task inputs, follow the source workflow and apply the explicit M11 qualifications below.

Requirements

Authorized platform exports/access and the upstream scoring, checks and platform skill packages.

Delivery and review notes

Published as attributed source text after an obvious-risk screen, not functional validation. Requires shared scoring-system.md, platform CHECKS/SKILL/REFERENCE packages and EXAMPLES.md, none closed/reviewed here. Universal 3x-CPA kill rule and GA4-as-arbiter assumptions are not generally valid evidence of wasted or incremental spend. Operational follow-up requires full dependencies and contextual attribution, lag and uncertainty handling; no ad-account changes or agents executed. Missing packages are not loaded automatically. Do not run installation commands, connect accounts or perform external writes merely because the original suggests them; resolve actual task requirements first.

Starting prompt

Use Cross-Platform Advertising Review for [BUSINESS/TASK]. Ask for missing inputs: Authorized platform exports/access and the upstream scoring, checks and platform skill packages. Return the source output format and label evidence, assumptions and validation steps. Published as attributed source text after an obvious-risk screen, not functional validation. Requires shared scoring-system.md, platform CHECKS/SKILL/REFERENCE packages and EXAMPLES.md, none closed/reviewed here. Universal 3x-CPA kill rule and GA4-as-arbiter assumptions are not generally valid evidence of wasted or incremental spend. Operational follow-up requires full dependencies and contextual attribution, lag and uncertainty handling; no ad-account changes or agents executed. Missing packages are not loaded automatically. Do not run installation commands, connect accounts or perform external writes merely because the original suggests them; resolve actual task requirements first.

Not for

Claiming the full orchestration ran when dependencies are unavailable or treating CPA thresholds as automatic stop orders.

What M11 added

M11 added task routing, German terms, scope and evidence requirements while preserving original authorship. Published as attributed source text after an obvious-risk screen, not functional validation. Requires shared scoring-system.md, platform CHECKS/SKILL/REFERENCE packages and EXAMPLES.md, none closed/reviewed here. Universal 3x-CPA kill rule and GA4-as-arbiter assumptions are not generally valid evidence of wasted or incremental spend. Operational follow-up requires full dependencies and contextual attribution, lag and uncertainty handling; no ad-account changes or agents executed. Missing packages are not loaded automatically. Do not run installation commands, connect accounts or perform external writes merely because the original suggests them; resolve actual task requirements first.

Original authorship remains with thatrebeccarae/claude-marketing · Original source ↗

Original license & copyright

MIT License

Copyright (c) 2026 Rebecca Rae Barton

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.

Source SHA-256: d8de92530075056f2824c64fe2e671356711b46e45c406b1d6411db3be598c02
Snapshot checked: 2026-09-28T15:02:41.083Z

Find a skill.Get to work.

All skills

Multi-Agent Orchestration

AI Context Window Audit

AI Agent Architecture Audit

AI Skill Discovery

Lead Magnet Strategy

Ideal Customer Profile

Go-to-Market Strategy

Marketing Campaign Ideas

Product Growth Loops

Competitor Analysis

North Star Metric

Product Positioning

Product Vision

Go-to-Market Motion Selection

Customer Feedback and JTBD Analysis

PESTLE Market Environment Analysis

Customer Journey Mapping

Ansoff Growth Options

Data Analysis Validation

Dataset Profiling

Statistical Analysis Guidance

Programmatic SEO Planning

Landing Page and Form Conversion Review

Paywall and Upgrade Planning

Signup and Registration Review

User Onboarding and Activation

Popup and Modal Planning

Email Sequence Copy and Flow

Marketing Campaign Planning

Marketing Content Drafting

Brand Voice and Content Review

Marketing Performance Reporting

Sales Company Research

Company and Contact Enrichment

Website Information Architecture

Content Strategy and Editorial Roadmap

Product Launch Planning

Customer Research and Voice of Customer

Community Growth and Member Experience

Competitor Research Profiles

Roadmap and Release Communication

Lifecycle Email Sequences

API Contracts and Interface Design

Observability and Instrumentation Planning

Agent Context and Session Handoff

Product Discovery Sprint

Deal Quality Scoring

Enrichment Waterfall Design

Customer Retention Playbook

Sales Coaching Practice

Revenue Cohort Analysis

Intent Signal Scoring

Customer Identity Matching

Segment Activation Planning

Sales Call Review

Retention Dashboard Design

Sales Coaching Competencies

Offer Delivery Design

Offer Value Communication

Service Productization

Conversion Hypothesis Prioritization

Braze Engagement Guidance

Brand Voice Guide Design

Content Pipeline Orchestration

Website Brand Profile

Buyer Objection Clarification

Klaviyo Integration Guidance

Offer Bonus Planning

Klaviyo Marketing Review

Cross-Platform Advertising Review

Expertise Business Model

LinkedIn Advertising Guidance

TikTok Advertising Guidance

Evidence Research Brief

Cinematic Video Prompts

Visual Prompt Model Adaptation

Visual Prompt Diagnostics

Short Film Development

Find a skill.
Get to work.