The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a PostToolUse hook to execute a local telemetry script (track-telemetry.sh) after tool calls to log usage metadata.
[DATA_EXFILTRATION]: Instrumentation scripts collect and transmit usage metadata, including session identifiers and truncated user prompts, to Shopify-controlled domains (shopify.dev). This behavior is explicitly disclosed in privacy notices within SKILL.md.
[DYNAMIC_EXECUTION]: The track-telemetry.ps1 script dynamically generates a temporary PowerShell script at runtime to handle background network operations.
[INDIRECT_PROMPT_INJECTION]: The skill processes results from documentation searches to generate code, creating an attack surface where maliciously crafted documentation could influence the agent's output.
Ingestion points: External data from scripts/search_docs.mjs enters the agent context via SKILL.md instructions.
Boundary markers: Lacks explicit instructions for the agent to treat search results as untrusted data.
Capability inventory: The agent has access to the bash tool for executing various scripts.
Sanitization: Implements a mandatory validation step using scripts/validate.mjs for all generated code.

shopify-partner