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.

B2B Outreach Foundation

Define the target account, prioritize buying signals, plan a human LinkedIn engagement routine and prepare evidence-led responses to buyer concerns.

Choose an area

01 · Category

What the MCP returns at this step

20 · Data & Analytics · Commerce Analytics

Shopify Revenue Reconciliation

Reconcile Shopify order reports with dashboard totals using dates, refunds, currencies and financial-status scope.

shopifyumsatzabgleichshopify-order-discipline

Shopify Order & Revenue Reconciliation · Original SKILL.md

---
name: shopify-order-discipline
description: "Use when querying Shopify orders or reconciling against a merchant dashboard. Default NO financial_status filter (dashboard parity), gross vs net revenue, created_at vs processed_at, refunds handling, multi-currency confirmation."
---
# Shopify Order Discipline

Before any Shopify order / revenue query goes to the user:

1. **Default to NO `financial_status` filter.** Shopify's "Total orders" / "Gross sales" tile counts ALL statuses (`pending`, `authorized`, `partially_paid`, `paid`, `partially_refunded`, `refunded`, `voided`). Adding `financial_status:paid` agent-side silently excludes pending + partially-paid orders and produces a number the merchant cannot reconcile against their dashboard.
2. **Clarify gross vs net BEFORE reconciling.** Shopify "Gross sales" = line items × qty (before discounts, refunds). "Net sales" = gross − discounts − refunds. 3rd-party MTAs (Cosmise, Triple Whale, Northbeam, etc.) typically report net. Never compare gross-against-net silently.
3. **Be explicit about timestamp field + timezone.** `created_at` (order created) vs `processed_at` (payment processed) can differ by hours. Orders placed near midnight UTC may roll to the next day in the merchant's local timezone.

---

## Rule 1 — No `financial_status` Filter By Default

| Status | Meaning | In "Total orders" tile |
|---|---|---|
| `pending` | Order placed, payment not captured | YES |
| `authorized` | Payment authorized, not captured | YES |
| `partially_paid` | Partial payment received | YES |
| `paid` | Full payment captured | YES |
| `partially_refunded` | Order kept; partial refund issued | YES (full original stays in gross) |
| `refunded` | Order fully refunded | YES (in gross; net subtracts the refund) |
| `voided` | Order voided | Typically excluded |

- Default Shopify orders query → NO `financial_status` filter. Result should match the merchant's "Total orders" / "Gross sales" tile.
- Add `financial_status:paid` ONLY when the user explicitly says "paid only" / "captured payments only" / "exclude pending".
- If the filter is set and the result doesn't match the dashboard, FIRST fix is to remove the filter and re-query — never invent "Meta + Google attribution overlap" or similar explanations.

**Forbidden** (without explicit user request): `query="created_at:>={start} AND created_at:<={end} financial_status:paid"` · any agent-authored exclusion of pending / partially-paid / refunded statuses.

---

## Rule 2 — Gross vs Net Revenue

| Concept | Definition | Where it shows up |
|---|---|---|
| **Gross sales** | line items × qty (before discounts, refunds) | "Gross sales" tile · API `total_line_items_price` |
| **Net sales** | gross − discounts − refunds | "Net sales" tile |
| **Total sales** | net + shipping + taxes | "Total sales" tile · API `total_price` |

3rd-party MTAs (Cosmise, Triple Whale, Northbeam, Wicked Reports, Lifetimely, Profitmetrics, Hyros, Rockerbox) typically report a NET-style figure.

- When the user says *"`{tool}` shows `{X}` for revenue"* and your number differs, FIRST hypothesis is gross-vs-net (or refund timing), not "tracking is broken".
- Never silently compare Shopify gross against a 3rd-party MTA net number — surface the methodology difference.
- State the revenue concept: *"Querying `total_price` for `{X}` gross; if `{tool}` shows net (after refunds), the gap is likely the partially-refunded portion."*

---

## Rule 3 — `created_at` vs `processed_at`

| Field | Meaning |
|---|---|
| `created_at` | When the order was created in Shopify |
| `processed_at` | When payment was processed / order moved out of pending |

Usually within minutes. Late-night UTC orders or delayed-capture gateways (Klarna, AfterPay batch overnight) can shift dates by a day.

- State the field + timezone: *"Querying `created_at:>={start} AND created_at:<={end}` in shop timezone `{tz}`."*
- If the dashboard total differs by a small count of orders, check whether the gap is `created_at` vs `processed_at` near the period boundary.
- When the merchant says "orders for `{month}`", confirm: orders CREATED in that month, or orders PROCESSED in that month?

---

## Rule 4 — Refund Handling

| Status | What happened | How to count |
|---|---|---|
| `partially_refunded` | Order kept, partial refund issued | Stays in count; gross stays in gross. Refunded $$ is in `refunds[].transactions[]` |
| `refunded` | Order fully refunded | Stays in count; gross stays in gross. Net subtracts entirely. |

- For gross, include all orders regardless of refund status.
- For net, subtract refund totals from gross. `refunds[].transactions[]` has per-order refund amounts.
- For partial refunds: deduct only the refunded portion, not the full order value. If your number is higher than the MTA's, partial refunds are the likely cause.

---

## Rule 5 — Multi-Currency

Shops can have a base `currency` AND orders denominated in other currencies. `total_price` is in the order's own currency; `total_price_set.shop_money.amount` is in the shop's base currency.

- Confirm the shop's base currency before reporting revenue. Don't assume USD.
- For consistency when summing across orders, use `total_price_set.shop_money.amount`.
- State the currency: *"All revenue figures in shop's base currency: `{CCY}`."*

---

## Rule 6 — User Disputes Your Shopify Number → Re-Query

When the user disputes your number with *"shopify said `{X}`"* / *"my dashboard shows `{X}`"* / *"in Shopify it's `{X}`"*:

1. Acknowledge the gap without defending the prior number.
2. Re-run with NO `financial_status` filter (Rule 1) and the right revenue field (Rule 2).
3. If the result now matches the dashboard: *"You're right — removing the `financial_status:paid` filter (which I had added) returns `{X}`, matching your dashboard."*
4. Do NOT blame "Meta + Google attribution overlap" or claim the dashboard is wrong.

---

## Examples

**Default order count for a window — user disputes**

❌ `query="created_at:>={start} AND created_at:<={end} financial_status:paid"` → `{N1}` orders. (User: *"shopify said `{N2}`"*.) Then: *"The gap is Meta + Google attribution overlap."*

✅ `query="created_at:>={start} AND created_at:<={end}"` (no `financial_status`) → `{N2}` orders. *"For `{period}` (shop tz `{tz}`): `{N2}` orders, `{gross}` gross sales. Includes pending, partially_paid, paid, partially_refunded — same scope as your Total Orders tile."*

**MTA shows lower gross than your Shopify number**

❌ *"4-hypothesis discrepancy: date logic, refund timing, status filter, timezone."* (Zero tool calls.)

✅ Re-pull Shopify with refund handling explicit. Then: *"Shopify net (gross minus refunds within the period): `{net}` — matches `{tool}`'s `{X}` within rounding. My earlier `{gross}` was gross (before refunds). The gap is `{N}` fully-refunded orders + the refunded portions of `{M}` partially-refunded orders."*

When to use

Reconcile Shopify order reports with dashboard totals using dates, refunds, currencies and financial-status scope.

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

Order data, dashboard definitions, date/timezone range, currencies and refund treatment.

Delivery and review notes

Orders may contain personal data: use task-scoped access and minimize retained/exported fields. Gross/net revenue, refunds and dates must be aligned before comparisons. 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 Shopify Revenue Reconciliation for [TASK]. Clarify Order data, dashboard definitions, date/timezone range, currencies and refund treatment. 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. Orders may contain personal data: use task-scoped access and minimize retained/exported fields. Gross/net revenue, refunds and dates must be aligned before comparisons. 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: aa76d01f86f480bb0926063e386dd83eaff997a93da8688d52feae6ee2088e85
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.