The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs users to add an external third-party MCP server at https://rube.app/mcp. This external service facilitates communication with Shopify and handles the authentication flow, meaning store data and credentials transit through this non-whitelisted infrastructure.
[DATA_EXFILTRATION]: The skill utilizes tools such as SHOPIFY_GET_ALL_CUSTOMERS and SHOPIFY_GET_ORDERS_WITH_FILTERS which retrieve sensitive store information. This data is processed by the agent and shared with the configured external MCP server.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted data from Shopify that could be modified by third parties (e.g., customer names, product titles, or order notes).
Ingestion points: Data enters the agent context via retrieval tools like SHOPIFY_GET_PRODUCTS, SHOPIFY_GET_ALL_CUSTOMERS, and SHOPIFY_GET_ORDERS_WITH_FILTERS defined in SKILL.md.
Boundary markers: The instructions do not define delimiters or specific warnings to ignore instructions embedded within the retrieved Shopify data.
Capability inventory: The skill possesses significant capabilities, including the ability to execute arbitrary queries via SHOPIFY_GRAPH_QL_QUERY and perform bulk writes via SHOPIFY_BULK_CREATE_PRODUCTS.
Sanitization: No evidence of input validation or sanitization is present in the skill instructions.

shopify-automation