n8n-expression-syntax

Pass

Audited by Gen Agent Trust Hub on Sep 7, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
  • [SAFE]: The skill is entirely instructional markdown and contains no executable scripts or command-line operations.
  • [DATA_EXPOSURE]: The skill documents the syntax for accessing environment variables (e.g., {{$env.API_KEY}}). This identifies access to sensitive data, but the documentation proactively mitigates risk by instructing users to use the platform's native credential system instead of expressions for sensitive keys.
  • [INDIRECT_PROMPT_INJECTION]: The skill provides instructions for ingesting untrusted data from webhooks (e.g., {{$json.body.name}}). While this describes an attack surface for indirect prompt injection, it also defines clear boundary markers (double curly braces) for data interpolation. Ingestion points: Webhook node output and $json.body (SKILL.md). Boundary markers: Explicit instructions to wrap dynamic content in double curly braces {{ }}. Capability inventory: Instructional only; documents access to previous node data, environment variables, and current JSON context. Sanitization: Not addressed in documentation, though it notes that certain contexts (like Code nodes) require direct variable access instead of template expressions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 7, 2026, 12:48 AM
Security Audit — agent-trust-hub — n8n-expression-syntax