Customer Intelligence Engine
Live customer data system that feeds every DTC Stack skill. Connects Shopify, Klaviyo, review platforms, CX tools, surveys, and analytics into one structured intelligence file.
- SKILL.MD FILE
- CONFIG TEMPLATES
- EXAMPLE OUTPUTS
- SETUP GUIDE
- INSTANT DELIVERY
What this skill does
Aggregates customer data from 6 source types into a central CUSTOMER_INTELLIGENCE.md file with 8 sections. Anomaly detection for week-over-week metric changes. On-demand, section-level, and scheduled refresh modes. Pre-Run integration blocks feed into all 19 downstream execution skills.
How to install
Download the skill package
Receive your files instantly after purchase.
Drop SKILL.md into directory
Move the file to your .claude/skills/ or equivalent agent directory.
Connect your Brand Brain
Reference your existing brain.yaml for on-brand outputs.
Start generating
Trigger the skill via command or chat interface.
Your AI skills work from stale data and assumptions - not what's actually happening in your store right now.
You run your email flow skill with last quarter's metrics. Your ad creative pulls from reviews you pasted in months ago. Your CRO audit doesn't know your top seller changed last week. Every skill produces decent output, but none of them know what's actually happening in your business today. You end up manually pasting data into every session, or worse, letting AI guess.
"Build me a post-purchase flow. My AOV is around $65 and I have about 30K subscribers. (Numbers from 3 months ago. Actual AOV is now $78 and you have 42K subscribers.)"
"Build me a post-purchase flow. (Skill reads live data: AOV is $78, 42K subscribers, top flow is Welcome at $4.20/recipient, repeat purchase rate is 31%, and customers complain most about shipping speed.)"
"Audit my store. Here's my URL. (Generic audit based on what the page looks like, no data on what's actually converting.)"
"Audit my store. (Skill reads live data: conversion rate dropped 12% this month, abandoned cart rate spiked on products over $80, top CX complaint is unclear sizing info, and review sentiment on your new product line is 3.2 stars vs 4.6 on your core line.)"
What you get.
6 Source Connectors
Shopify, Klaviyo, review platforms (Junip, Yotpo, Judge.me, Stamped, Okendo), CX platforms (Gorgias, Richpanel, Commslayer, Zendesk), surveys (Typeform, Fairing), and analytics (GA4, Triple Whale). MCP-first, REST API fallback, manual paste as last resort.
Central Intelligence File
One CUSTOMER_INTELLIGENCE.md with 8 sections: store performance, email/SMS intel, voice of customer, product intelligence, retention data, trends, and gaps. Every downstream skill reads from it automatically.
Anomaly Detection
Flags any metric that changes more than 15% week-over-week. Revenue dropped? Review sentiment shifted? Churn spiked? You see it in the Intelligence Summary before it becomes a problem.
On-Demand + Scheduled Refresh
Refresh whenever you want, or set up automatic weekly pulls. Section-level refresh when you only need fresh data for one area. Works with 1 source minimum - the more you connect, the smarter every skill gets.
Connects to your data. Every skill gets smarter.
One connection feeds live data straight into this skill. Use one source, all of them, or none - it works either way.
Shopify
EcommerceProvides: Revenue, AOV, conversion rate, inventory signals, and post-purchase survey data.
Why it matters: The intelligence file grounds every skill in your actual store performance, not last quarter's memory.
Klaviyo
Email & SMSProvides: Flow performance, campaign revenue, list growth, and segment behavior.
Why it matters: Email decisions across all skills reference what is actually happening on your list this week.
Any one of these Attribution sources:
Polar Analytics
AttributionProvides: Blended CAC, MER, ROAS by channel, and cohort retention.
Why it matters: Every skill reads the same attribution source of truth, no spreadsheet reconciliation.
Kleio
AttributionProvides: Blended CAC, MER, ROAS by channel, and cohort retention.
Why it matters: Every skill reads the same attribution source of truth, no spreadsheet reconciliation.
GA4
AttributionProvides: Traffic by channel, landing page behavior, and assisted conversions.
Why it matters: Organic and on-site behavior data lands in the intelligence file alongside paid data.
Any one of these Reviews sources:
Judge.me
ReviewsProvides: Review sentiment trends, verbatim VOC, and product-level praise or complaint patterns.
Why it matters: Voice-of-customer signals refresh weekly, not once a year in a research sprint.
Junip
ReviewsProvides: Review sentiment trends, verbatim VOC, and product-level praise or complaint patterns.
Why it matters: Voice-of-customer signals refresh weekly, not once a year in a research sprint.
Yotpo
ReviewsProvides: Review sentiment trends, verbatim VOC, and product-level praise or complaint patterns.
Why it matters: Voice-of-customer signals refresh weekly, not once a year in a research sprint.
Any one of these CX sources:
Gorgias
CXProvides: Pre-sale question themes, complaint patterns, and cancellation reasons.
Why it matters: Every skill sees what is actually causing churn and friction this month.
Commslayer
CXProvides: Pre-sale question themes, complaint patterns, and cancellation reasons.
Why it matters: Every skill sees what is actually causing churn and friction this month.
Richpanel
CXProvides: Pre-sale question themes, complaint patterns, and cancellation reasons.
Why it matters: Every skill sees what is actually causing churn and friction this month.
Any one of these Subscriptions sources:
Recharge
SubscriptionsProvides: Subscription status, billing events, swap and pause history.
Why it matters: Retention and win-back skills read subscription lifecycle data, not just one-time orders.

Skio
SubscriptionsProvides: Subscription contracts, billing events, and customer preferences.
Why it matters: Retention and win-back skills read subscription lifecycle data, not just one-time orders.
Levanta
AffiliateProvides: Partner list, product catalog, and affiliate link performance.
Why it matters: Affiliate and partnership signals land alongside your owned-channel data.
Runs without integrations - intelligence file uses manual inputs and Shopify-native data only. Add connections as you grow.
When to use this.
Setting up a new brand and pulling all available data into one intelligence file before running any skills
Weekly refresh to keep all downstream skills working from current data instead of assumptions
Diagnosing why a metric changed - the anomaly detection cross-references data across all sources
Running a targeted section refresh before a specific skill (e.g., refresh email data before building Klaviyo flows)
Common questions.
Is this the same as the Brand Brain?
No. The Brand Brain (Commerce Intelligence System) captures who you are - voice, positioning, personas, objections. The Customer Intelligence Engine captures what's happening - live metrics, customer feedback, trends. Skills read both. Together, they give AI complete context.
How many sources do I need connected?
One is enough to start. Shopify alone gives you store performance and product intelligence. Add Klaviyo for email data, a review platform for voice of customer, and so on. Each source makes the intelligence richer.
How does it connect to my platforms?
MCP via Composio is the recommended method - install once, authenticate via browser, no API keys to manage. REST API is the fallback for platforms without MCP support. Manual paste works for everything else.
Will this slow down my other skills?
No. If the intelligence file exists, skills read it for context. If it doesn't exist or is stale, skills run normally with Brand Brain context only. It's a context enhancer, not a gate.
How often should I refresh?
Weekly is the sweet spot for most brands. More often during high-velocity periods like launches or big sales. The skill flags when data is getting stale.
Stop pasting stale data into every session.
Connect your store, email, reviews, and CX platform once. Every skill reads live data automatically.
$199 · ALL 20 SKILLS · INSTANT DELIVERY
Want us to build it for you?