n8n-workflow-builder
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely instructional and operates by generating configuration data (JSON) rather than executing code or scripts. It does not possess any native capabilities for network access, file system modification, or credential harvesting.
- [PROMPT_INJECTION]: The skill defines a workflow for processing untrusted user input (such as diagrams or descriptions) to generate functional automation logic. This is an indirect prompt injection surface, as malicious instructions could be embedded in the input to influence the generated JSON. However, the risk is mitigated by the requirement for a human to manually import the JSON and explicitly configure credentials within their own n8n environment.
- Ingestion points: User-provided text, diagrams, and screenshots (SKILL.md Phase 1).
- Boundary markers: Absent; there are no explicit delimiters instructed for separating user input from the builder logic.
- Capability inventory: The generated JSON can specify nodes for network requests (HTTP), CRM access (HubSpot, Salesforce), and messaging (Slack), though these are not executed by the skill itself.
- Sanitization: None; the skill relies on the LLM to correctly map user intent to valid n8n node parameters.
Audit Metadata