hydrogen-analytics-tracking

Pass

Audited by Gen Agent Trust Hub on May 28, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs users to load tracking scripts from well-known and trusted services, specifically Google Tag Manager (googletagmanager.com) and Google Analytics (google-analytics.com). These are standard practices for analytics implementation and are considered safe.
  • [DATA_EXFILTRATION]: The core functionality involves sending tracking data (events, product views, purchases) to third-party vendors such as Google, Meta, and TikTok. The skill explicitly mandates server-side hashing for sensitive user data like email and phone numbers before transmission, which is a security best practice for data privacy.
  • [COMMAND_EXECUTION]: The skill mentions local maintenance scripts (scripts/search_shopify_docs.mjs) for searching documentation. While the code for these scripts is not provided, their described use case is benign.
  • [PROMPT_INJECTION]: No patterns of prompt injection, role-play overrides, or system prompt extraction instructions were detected in the skill instructions or reference files.
  • [CREDENTIALS_UNSAFE]: The skill uses appropriate placeholders for secrets (e.g., GA4_API_SECRET, META_CAPI_ACCESS_TOKEN, SHOPIFY_WEBHOOK_SECRET) and provides clear instructions in gotchas.md for secret rotation and management, discouraging the storage of clear-text credentials.
  • [INDIRECT_PROMPT_INJECTION]: The skill creates an attack surface by ingesting data from external sources (Shopify webhooks and browser-side sendBeacon requests). However, it mitigates this risk by recommending strict schema validation and server-side processing, and it lacks dangerous capabilities like arbitrary code execution that would make this surface highly exploitable.
  • Ingestion points: Webhook endpoint /api/webhooks-orders-create.ts and tracking API /api/track.ts.
  • Boundary markers: Encourages JSON schema validation (validate(raw)) as seen in architecture.md.
  • Capability inventory: Capabilities are limited to outgoing network requests to analytics vendors.
  • Sanitization: Implements server-side PII hashing as described in architecture.md.
Audit Metadata
Risk Level
SAFE
Analyzed
May 28, 2026, 08:54 PM
Security Audit — agent-trust-hub — hydrogen-analytics-tracking