Partial Execution for AI Tools (HTTP tool)

Feature Requests
1

The idea is:

Partial Execution for AI Tools is a great feature to reduce costs and latency during testing. However, since the popup relies on the $fromAI("message") function, it’s currently not compatible with the HTTP Request tool, which requires using the {message} syntax instead.

Two possible improvements (in order of preference):

  1. Enable support for $fromAI("message") in the HTTP Request tool.
  2. Extend the popup’s compatibility to also recognize {variable} syntax, not just $fromAI("...").

My use case:

I often rely on the HTTP node either because:

  • The desired tool isn’t supported natively.
  • The native node is not fully implemented or lacks certain features.

Why this matters:

Making the HTTP node compatible with Partial Execution would bring it in line with other nodes, allowing for more consistent and flexible development of AI-based workflows.

2

I found a comment published in March from a N8N member here that mentioned the ability to support $fromAI in the HTTP tool. Keep posted !