The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill provides instructions for creating n8n workflows that process data from external, potentially untrusted sources such as webhooks and HTTP requests, creating a surface for indirect prompt injection.
Ingestion points: SKILL.md identifies n8n-nodes-base.webhook and n8n-nodes-base.httpRequest as primary nodes for bringing external data into a workflow.
Boundary markers: The documented workflow JSON patterns do not include delimiters or instructions to treat external data as untrusted content.
Capability inventory: The skill documents capabilities for performing network requests (httpRequest node) and executing custom JavaScript (code node) within the automation environment.
Sanitization: There are no provided examples of input validation or sanitization for dynamic expressions like ={{ $json.field }} used to interpolate external data.

n8n