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

66 · Marketing · Creative Research

Competitor Ad Collection

Collect competitor ad samples and separate long-running from recently repeated creatives. Use Competitor Ad Intelligence for a cross-brand report.

wettbewerbsanzeigensammlungcreative-research

Competitor Creative Research · Original SKILL.md

---
name: creative-research
description: "Use when collecting live competitor ads and classifying evergreen (long-running) versus breakout (recently repeated) investment signals with GoMarble Ads Library tools. For a summarized cross-competitor teardown and whitespace report, use competitor-ad-intelligence."
---
# Creative Competitor Research Protocol

Discover new creative concepts by analyzing competitor ads. Use observable longevity and variation signals to prioritize concepts, then adapt them for your brand without treating those proxies as verified performance.

## Prerequisites

Before exploring competitors, identify your current winning ads and their patterns (angles, formats, hooks). This gives you a baseline to compare competitor insights against. Use your primary conversion metric: Purchase ROAS (ecommerce) or Cost Per Result (leadgen).

---

## Step 1: Find Competitors

### AI-Powered Discovery

**Tool: `ads_library_find_competitors`**
```
Parameters:
website_url: "www.yourcompany.com" # Preferred
OR brand: "Your Brand Name"
```
Returns: competitor brands (up to 5) with names + URLs, product keywords, niche category, market target (B2B/B2C), suggested search queries.

### Competitor Selection Rules

- Match competitors by: same product category, similar AOV/price point, similar market size
- A small DTC brand can't replicate a $100M+ brand's approach — scale matters
- Search "[category]" and see which brands appear in results
- DO NOT use generic/wide search terms (e.g., "shopify apps"). Use specific category descriptors (e.g., "sale countdown shopify apps")

### When AI Discovery Returns Zero Ads

1. Define 2–3 parallel categories (NOT wider, but adjacent). Example: "sale countdown shopify apps" → try "conversion boost shopify apps", "sale announcement shopify apps"
2. For each parallel category, find the biggest brands first
3. Try `ads_library_search_brands` with those brands BEFORE using keyword search
4. Only use `ads_library_discover_ads` keyword search as last resort
5. NEVER go generic. Search terms must be relevant to the specific product function.

---

## Step 2: Get Brand IDs

**Tool: `ads_library_search_brands`**
```
Parameters:
domain: "competitor.com" # More accurate than name
OR name: "Competitor Brand"
limit: 10
```
Returns: brand_id (needed for fetching ads), brand name, category/niches. Save brand_ids for Steps 3A/3B.

---

## Step 3A: Evergreen Signals (Sustained Investment)

**Goal**: Find ads running 30+ days. Longevity indicates continued investment, not verified profitability.

**Tool: `ads_library_get_ads_by_brand_id`**
```
Parameters:
brand_ids: ["brand_id_1", "brand_id_2"]
live: "true"
order: "longest_running"
limit: 50
running_duration_min_days: 30 # Optional filter
```

### Analysis Process

1. Filter: remove ads tagged "low impression count". Focus on top 10–15 longest runners, prioritize 60+ days.
2. Deep analyze with **`ads_library_analyze_ad`**: pass top ad IDs with `include_duplicates: true`.
3. Document patterns across the prioritized ads:

| Element | What to Look For |
|---------|-----------------|
| **Angle** | What selling points repeat across prioritized ads? (Authority? Results? Price?) |
| **Concept** | How is the angle executed? (Testimonial? Before/after? Expert?) |
| **Visual Style** | Professional vs lo-fi? UGC? Product-focused? |
| **Script** | Common structure? (Problem → Solution → CTA) Length patterns? |
| **Hook** | Visual patterns? Text overlay? Audio approach? (first 3–5 seconds) |
| **Format** | Dominant format? (Testimonial, demo, lifestyle) |

4. For each prioritized concept, suggest adaptation: their approach → your version with rationale.

---

## Step 3B: Breakout Signals (Recent Repetition)

**Goal**: Find concepts launched recently and repeated aggressively. This indicates repeated investment or structured testing, not confirmed scaling success.

**Tool: `ads_library_get_ads_by_brand_id`**
```
Parameters:
brand_ids: ["brand_id_1", "brand_id_2"]
live: "true"
start_date: "[30 days ago, YYYY-MM-DD]"
order: "newest"
limit: 100
```

### Identifying Breakout Candidates

Group ads by similarity (same copy/format/visual style). A Breakout Candidate is a concept that:
- Launched within 30 days
- Has been repeated/varied multiple times
- Variations launched across different dates (not just one batch)

**Repetition signal levels**:
- **HIGH** (4+ variations across multiple dates): repeated investment; prioritize for analysis
- **MEDIUM** (2–3 variations): Worth monitoring
- **LOW** (< 2): Testing phase

Deep analyze with `ads_library_analyze_ad` using `include_duplicates: true`.

---

## Step 4: Alternative Discovery

### By Keyword/Product

**Tool: `ads_library_discover_ads`**
```
Parameters:
search_query: "protein shake" # Max 6 words, specific not generic
order: "longest_running"
niches: ["health/wellness"]
display_format: ["video"]
live: true
limit: 50
```
Use when researching a category broadly rather than specific brands.

### Brand Analytics

**Tool: `ads_library_get_brand_analytics`**
```
Parameters:
brand_ids: ["brand_id_1"]
```
Returns: total active ads, format distribution, platform breakdown. Use to decide which competitors are most active and worth deep analysis.

---

## Key Principles

1. **Scale matters** — compare similar-sized competitors
2. **Patterns > individual ads** — look for what repeats across prioritized ads
3. **Recent and old answer different questions** — recent repetition shows current testing/investment; longevity shows sustained investment
4. **Variations are a proxy** — multiple versions signal testing or investment, not verified performance
5. **Adaptation not copying** — translate concepts authentically to your brand
6. **Recency matters** — current repetition can justify prioritizing a timely, controlled test; it does not prove the concept is scaling successfully

When to use

Collect competitor ad samples and separate long-running from recently repeated creatives. Use Competitor Ad Intelligence for a cross-brand report.

What you get

The complete original GoMarble guidance, with source attribution and declared limits.

How it works

Clarify scope, use available evidence, and apply the relevant source workflow within actual user authorization.

Requirements

Competitor brands, market, date window and authorized ad-library access.

Delivery and review notes

Ad longevity and repeated variants are investment proxies, not verified competitor spend, profitability or causal success. GoMarble connectors, referenced sibling skills and runtime dependencies are not bundled or installed by M11. Brief obvious-danger screening only; no functional test or comprehensive safety certification. Loading this text authorizes no external calls, account mutations or ad spend.

Starting prompt

Use Competitor Ad Collection for [TASK]. Clarify Competitor brands, market, date window and authorized ad-library access. Separate observed evidence from heuristics and recommendations. Check actual connector availability and approval semantics before any external action.

Not for

Claiming installed GoMarble tools, proven campaign outcomes or authorization to launch ads merely from loading this source.

What M11 added

German routing and explicit scope. Original attribution: GoMarble. Ad longevity and repeated variants are investment proxies, not verified competitor spend, profitability or causal success. GoMarble connectors, referenced sibling skills and runtime dependencies are not bundled or installed by M11. Brief obvious-danger screening only; no functional test or comprehensive safety certification. Loading this text authorizes no external calls, account mutations or ad spend.

Original authorship remains with gomarble-ai/marketing-agent · Original source ↗

Original license & copyright

MIT License

Copyright (c) 2026 GoMarble

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: df6aca22b7d9ccec6cedafc59964b4153faacee403bb5c5a4770e9f2163a0e9a
Snapshot checked: 2026-09-28T16:19:52.867Z

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.