The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill includes patterns for processing data retrieved from the Shopify GraphQL and Bulk Operations APIs. This creates an attack surface where data managed within a Shopify store (such as product descriptions or customer metadata) could contain instructions intended for the AI agent.
Ingestion points: The paginateShopify generator in references/cursor-pagination.md and the runBulkOperation function in SKILL.md both ingest external API data.
Boundary markers: The provided code snippets do not include explicit delimiter boundaries or instructions to the agent to ignore embedded commands in the data.
Capability inventory: The skill utilizes tools for reading/writing files and making network requests to Shopify APIs.
Sanitization: The ingested data is parsed as JSON/JSONL but is not sanitized for potential prompt injection patterns before being yielded to the application logic.
[COMMAND_EXECUTION]: The documentation suggests running npx graphql-codegen to synchronize TypeScript types with the Shopify schema. This is a standard and safe developer workflow for projects using GraphQL and code generation.
[EXTERNAL_DOWNLOADS]: The bulk operation helper in SKILL.md performs a network fetch to a URL returned by the Shopify API. This is the documented and necessary method for retrieving bulk data exports from Shopify's infrastructure.
[CREDENTIALS_SAFE]: The implementation in references/typed-graphql-client.md correctly retrieves sensitive credentials like SHOPIFY_API_SECRET and SHOPIFY_ACCESS_TOKEN from environment variables, avoiding the risk of hardcoded secrets.

shopify-sdk-patterns