Skills
Navigate
skills/display-design-studio/skills/shopify-polaris-checkout-extensions/Gen Agent Trust Hub

shopify-polaris-checkout-extensions

Warn

Audited by Gen Agent Trust Hub on Jul 31, 2026

Risk Level: MEDIUMDATA_EXFILTRATIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTIONCREDENTIALS_UNSAFE
Full Analysis
  • [METADATA_POISONING]: The skill's frontmatter identifies the author as "Shopify", which contradicts the author context of "display-design-studio".
  • [DATA_EXFILTRATION]: Telemetry hooks in scripts/track-telemetry.sh and scripts/track-telemetry.ps1 are configured to capture the user's prompt verbatim and stash it in a temporary file ($TMPDIR/shopify-ai-toolkit-telemetry-...). This stashed data is subsequently sent to https://shopify.dev/mcp/usage during tool usage events.
  • [COMMAND_EXECUTION]: The skill mandates the execution of scripts/search_docs.mjs and scripts/validate.mjs for every response. These scripts involve local file parsing and network communication.
  • [EXTERNAL_DOWNLOADS]: scripts/search_docs.mjs and scripts/validate.mjs perform network requests to shopify.dev to fetch documentation and usage information.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user content through its validation pipeline.
  • Ingestion points: scripts/validate.mjs (via CLI arguments).
  • Boundary markers: Absent.
  • Capability inventory: Performs network operations and script execution (via node/bash).
  • Sanitization: Uses JSON.stringify for telemetry reporting.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 31, 2026, 09:26 AM
Security Audit — agent-trust-hub — shopify-polaris-checkout-extensions