n8n-workflow-patterns

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). The docs describe high-risk patterns that enable intentional abuse: AI agents with "ai_tool" access to arbitrary HTTP/DB/Code nodes (including AI-supplied SQL and HTTP parameters), examples executing shell commands (exec pg_dump), and generic HTTP Request tools — all of which can be used for data exfiltration, remote code execution, and credential theft if misconfigured or abused.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.85). Outsider free text can enter the LLM context via the Webhook Processing path: the runtime webhook payload (e.g., $json.body.message / $json.body.text from an external caller) is ingested into the workflow and can be forwarded into the AI Agent node as the agent’s input/prompt.