Improving $fromAI Parameter Handling: Reducing Duplication and Enabling Enums for Better LLM Integration

This feature request reveals a common integration gap between low-code automation platforms and LLM tool schemas. When $fromAI is used, missing parameter descriptions and absent enums force developers to duplicate documentation and deliver less-structured inputs to models, which increases hallucination risk and reduces predictable behavior. Ensuring that the runtime tool schema exposes parameter descriptions and static option sets as enums would provide stronger signals to models and a cleaner experience for users.

The business significance is clear: better schema fidelity improves both developer productivity and runtime reliability. For teams building enterprise automations, consistent, explicit parameter contracts reduce debugging time, make prompts more deterministic, and lower the cost of maintaining automations across model upgrades. Enabling enums can also allow models to choose from constrained sets, reducing invalid outputs and downstream error handling.

From an implementation standpoint, n8n (or similar platforms) should adopt two changes: default to using parameter.description when $fromAI would otherwise be empty, and translate static options / resourceLocator / listSearch results into enum or oneOf definitions in the runtime tool schema. Additional improvements include versioning the runtime schema, surfacing which fields are model-populated, and allowing optional developer overrides for enriched descriptions.

Actionable recommendations for leaders evaluating automation platforms: prioritize vendors that expose rich tool schemas and LLM integration controls, require auditability for model-generated parameters, and push for platform features that reduce duplicated documentation. Internally, enforce prompt engineering standards that rely on platform-provided metadata to keep prompts concise and reliable - this lowers maintenance overhead and increases predictability in production automations.