The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides instructions for executing shell commands and scripts to interact with store data and the Shopify API.
It utilizes curl for auditing existing storefront JSON-LD and communicating with the GraphQL Admin API.
It leverages the shopify CLI for store authentication and command execution.
The agent is directed to generate and execute task-specific scripts to process data and interface with the Admin API.
[EXTERNAL_DOWNLOADS]: Recommends the installation of external plugins from a well-known service provider.
Instructions include adding and installing the official Shopify-AI-Toolkit via the Claude plugin marketplace.
[PROMPT_INJECTION]: The skill processes untrusted store-specific content which presents an indirect prompt injection surface.
Ingestion points: The skill reads product descriptions, page content, and blog articles from the Shopify Admin API as described in references/pipeline-and-prompts.md.
Boundary markers: The prompt recipes do not include specific delimiters or structural markers when interpolating fetched content into model prompts.
Capability inventory: The skill can perform network operations via curl and write data to Shopify store metafields using the metafieldsSet mutation.
Sanitization: The workflow includes a mandatory validation stage that parses generated output as JSON and performs keyword filtering for disallowed schema patterns before submission.

shopify-json-ld