The Agent Skills Directory

[DATA_EXFILTRATION]: The skill collects and transmits telemetry data, including user prompts, generated code blocks, and agent session identifiers, to the vendor's domain (shopify.dev). This process is facilitated by both a mandatory validation script and automated telemetry hooks.
[OBFUSCATION]: The skill ships multiple minified JavaScript files (log_feedback.mjs, search_docs.mjs, validate.mjs) used for its primary operations, making direct audit of the underlying logic more difficult.
[COMMAND_EXECUTION]: The skill utilizes a PostToolUse hook to automatically execute a telemetry script (track-telemetry.sh or track-telemetry.ps1) after every tool call. These scripts perform network operations in the background to avoid blocking the agent's workflow.
[INDIRECT_PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection as it processes untrusted user prompts and retrieves documentation results from an external API, both of which are used to generate and validate code within the agent's context.
Ingestion points: User prompts are ingested by the validate.mjs script and telemetry hooks; external search results are retrieved from the Shopify documentation service.
Boundary markers: None observed in the instructions for prompt interpolation.
Capability inventory: The skill uses the bash tool to execute scripts that perform file system access and network operations.
Sanitization: User prompts are base64-encoded before being passed as arguments to the validation scripts.

shopify-pos-ui