shopify-admin-agentic-…
Edit the theme's robots.txt.liquid to explicitly allow AI crawlers (GPTBot, ClaudeBot,…
Create a formal Shopify Return record for an order, specifying line items, quantities, and return reason — the first step in the native returns workflow.
$ npx -y skills add 40rty-ai/shopify-admin-skills --skill shopify-admin-return-initiation --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/shopify-admin-return-initiationContext preview
The summary Claude sees to decide when to auto-load this skill.
Create a formal Shopify Return record for an order, specifying line items, quantities, and return reason — the first step in the native returns workflow.
name: shopify-admin-return-initiation role: customer-support description: "Create a formal Shopify Return record for an order, specifying line items, quantities, and return reason — the first step in the native returns workflow." toolkit: shopify-admin, shopify-admin-execution api_version: "2025-01" graphql_operations: - order:query - returnCreate:mutation status: stable compatibility: Claude Code, Cursor, Codex, Gemini CLI
Initiates a formal Shopify Return on a delivered order — specifying which line items to return, quantities, and reason. This creates the return record in Shopify's native returns system (distinct from simply issuing a refund). Used by support agents when a customer contacts them to return delivered items. The return record enables tracking, warehouse inspection, and exchange/refund resolution downstream. Note: `returnCreate` requires the order to be in `FULFILLED` status. For orders that haven't shipped yet, use `cancel-and-restock` instead. For already-returned items needing a refund, use `refund-and-reorder`.
| Parameter | Type | Required | Default | Description | |-----------|------|----------|---------|-------------| | store | string | yes | — | Store domain (e.g., mystore.myshopify.com) | | format | string | no | human | Output format: `human` or `json` | | dry_run | bool | no | false | Preview operations without executing mutations | | order_id | string | yes | — | GID of the order (e.g., `gid://shopify/Order/12345`) | | return_line_items | array | no | all fulfilled | Array of `{fulfillment_line_item_id, quantity, reason, reason_note}` to return | | return_reason | string | no | `OTHER` | Default return reason for all items if not specified per-item: `SIZE_TOO_SMALL`, `SIZE_TOO_LARGE`, `WRONG_ITEM`, `NOT_AS_DESCRIBED`, `DEFECTIVE`, `STYLE`, `COLOR`, `UNWANTED`, `OTHER` | | notify_customer | bool | no | true | Send return initiation notification email to customer |
> ⚠️ Step 2 executes `returnCreate` which creates a formal return record and — if `notify_customer: true` — sends an email to the customer. This is appropriate only after verifying with the customer that a return is expected. Run with `dry_run: true` to preview the return line items and quantities before committing.
1. **OPERATION:** `order` — query **Inputs:** `id: <order_id>` **Expected output:** Order `name`, `displayFulfillmentStatus` (must be `FULFILLED` — abort if not), `fulfillments` with `fulfillmentLineItems` including `id`, `quantity`, `discountedTotalSet`
2. **OPERATION:** `returnCreate` — mutation **Inputs:** `returnInput.orderId`, `returnInput.returnLineItems` array (each with `fulfillmentLineItemId`, `quantity`, `reason`, `customerNote`) **Expected output:** `return.id`, `return.status: OPEN`, `userErrors`
# order:query — validated against api_version 2025-01
query OrderForReturn($id: ID!) {
order(id: $id) {
id
name
displayFulfillmentStatus
displayFinancialStatus
customer {
id
defaultEmailAddress {
emailAddress
}
firstName
lastName
}
fulfillments {
id
status
fulfillmentLineItems(first: 50) {
edges {
node {
id
quantity
lineItem {
title
variant {
id
sku
}
}
discountedTotalSet {
shopMoney { amount currencyCode }
}
}
}
}
}
}
}# returnCreate:mutation — validated against api_version 2025-01
mutation ReturnCreate($returnInput: ReturnInput!) {
returnCreate(returnInput: $returnInput) {
return {
id
status
order {
id
name
}
}
userErrors {
field
message
}
}
}**Claude MUST emit the following output at each stage. This is mandatory.**
**On start**, emit:
╔══════════════════════════════════════════════╗ ║ SKILL: return-initiation ║ ║ Store: <store domain> ║ ║ Started: <YYYY-MM-DD HH:MM UTC> ║ ╚══════════════════════════════════════════════╝
**After each step**, emit:
[N/TOTAL] <QUERY|MUTATION> <OperationName>
→ Params: <brief summary of key inputs>
→ Result: <count or outcome>If `dry_run: true`, prefix every mutation step with `[DRY RUN]` and do not execute it.
**On completion**, emit:
For `format: human` (default):
══════════════════════════════════════════════ OUTCOME SUMMARY Order: <name> Return ID: <id> Line items included: <count> Return status: <status> Customer notified: <yes/no> Errors: 0 Output: none ══════════════════════════════════════════════
For `format: json`, emit:
{
"skill": "return-initiation",
"store": "<domain>",
"started_at": "<ISO8601>",
"completed_at": "<ISO8601>",
"dry_run": false,
"steps": [
{ "step": 1, "operation": "OrderForReturn", "type": "query", "params_summary": "order <id>", "result_summary": "<status>", "skipped": false },
{ "step": 2, "operation": "ReturnCreate", "type": "mutation", "params_summary": "<n> line items, reason: <reason>", "result_summary": "return <id>", "skipped": false }
],
"outcome": {
"order_name": "<name>",
"return_id": "<id>",
"return_status": "<status>",
"line_items_count": "<count>",
"customer_notified": "<true/false>",
"errors": 0,
"output_file": null
}
}No CSV output. The session summary reports the return ID and status. Line items included in the return are listed in the step 2 output.
| Error | Cause | Recovery | |-------|-------|----------| |
Community-maintained AI agent skills for operating Shopify stores — workflows, optimization, reports and more
Edit the theme's robots.txt.liquid to explicitly allow AI crawlers (GPTBot, ClaudeBot,…
Rewrite thin product descriptions into structured, fact-rich copy (materials, fit, use-cases,…
Generate and set descriptive alt text on product images so AI agents (which can't 'see'…
Generate and publish an /llms.txt guide (brand summary, flagship products, key policies,…
Define and populate agentic-commerce metafields (material, attributes, key features, specs,…
Inject an Organization JSON-LD block (name, logo, sameAs social links, contactPoint) into the…