gtm-implementation
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from project-level configuration files to perform automated code modifications and API operations.
- Ingestion points: The skill reads
gtm-tracking-plan.json(containing events and parameters) andgtm-context.md(containing business logic context) to determine implementation steps. - Boundary markers: There are no explicit delimiters or instructions provided to the agent to disregard malicious commands that might be embedded within these external data files.
- Capability inventory: The skill has high-privilege capabilities including the ability to modify project source code (e.g., React/Next.js/Vue components) via the
Edittool and the ability to perform network operations via the Google Tag Manager API. - Sanitization: The logic does not show validation or escaping of the strings extracted from the tracking plan before they are interpolated into
window.dataLayer.pushcalls or API request payloads.
Audit Metadata