IF Display Omni🔍

IF Display Omni is the pack's "show me what's actually in this wire" node, built for the weirdest type in ComfyUI: the OMNI. The IF_AI pack defines OMNI as a universal wildcard that matches anything - it's how IF Chat Prompt's Tool_Output can carry a string, a dict, or an Omost canvas layout without knowing ahead of time which it'll be. That's flexible, but it's also opaque. This node takes any OMNI input, inspects it, prints it to the console, and - this is the part that matters - separates out Omost canvas conditionings so they can flow into an Omost-capable sampler.

Think of it as the toolbelt between the chat node and the image side of an Omost workflow. When IF_ChatPrompt runs with the omost_tool enabled, the Tool_Output is a bundle of canvas regions and their prompts. IF Display Omni extracts those conditionings and hands them out the canvas_conditioning output, while the raw text rides along in text_output. Without it, that payload is stuck inside a black-box port.

How it works

The node accepts omni_input and runs it through a small classifier at execution time. If a value is a dict containing conditionings, it treats it as batched canvas data and extracts it. If it's a list of dicts, it forwards that directly as a canvas-conditioning list. If it's a plain string, it just passes the text through. Anything else gets JSON-serialized for display. Whatever it finds, it pushes into the node's UI so you can see it, and updates the workflow's stored widget values.

The outputs are canvas_conditioning (type OMOST_CANVAS_CONDITIONING - feed this to an Omost canvas node/sampler) and text_output (a string with the human-readable content).

The inputs that matter

• omni_input - the OMNI wire from an IF node's tool output. That's the entire input surface, and that's fine; the node's job is output.

Installation

Pack standard:

cd ComfyUI/custom_nodes
git clone https://github.com/if-ai/ComfyUI-IF_AI_tools.git
pip install -r requirements.txt

Or "IF_AI_tools" via ComfyUI Manager and restart.

Gotchas

Two things worth flagging. First, the Omost side of this pack needs the author's tuned model to be worthwhile - ollama run impactframes/dolphin_llama3_omost is the README's recommendation, and the plain-text route through a normal model won't produce useful canvas data. Second, it's honestly a niche node: if you're not doing Omost layout workflows, you'll use it maybe once to debug what a tool returned, then forget it's there. That's fine - it's a debug lens with one real job, and it does that job. The pack is archived and Omost tooling has since moved into the comfy-deploy LLM toolkit, but the canvas-conditioning handoff pattern lives on there.