LoRA Stack Info Display

You know the drill: a workflow loaded five LoRAs, and you have no idea what any of them actually do, what their trigger words are, or what weight you left them at. LoRA Stack Info Display is the node that answers all of that at a glance - it takes a LoRA stack and renders a formatted markdown report for every entry, complete with Civitai details, weights, and links.

The input is a single LORA_STACK - the same stack type that Simple Lora Stack, Triple Lora Stack, and the other Sage Utils LoRA nodes produce and consume. Feed the stack's output into this node and it walks each entry, pulls the LoRA's name and its model/clip weights, computes the file's hash, and looks the version up in the pack's Civitai cache. What comes back on markdown_display is a structured summary per LoRA: model name, version, base model, the weight you applied, and a direct link to the Civitai page. The whole thing is one markdown string that renders in ComfyUI's text display, so it doubles as a nice note-to-self in a shared workflow.

How does it know the Civitai details? This is where the pack's machinery kicks in. Sage Utils caches Civitai model info locally in comfyui/user/default/SageUtils/ as sage_cache_hash.json and sage_cache_info.json - first lookup for a LoRA hits the API and populates the cache, later lookups are fast and offline-friendly. LoRAs you've loaded before will show rich info instantly; one you've never touched may show less until it's been hashed and fetched. It's the same cache that powers the pack's keyword extraction and model-scan nodes, so the more you use Sage Utils, the better this display gets.

The honest caveat: this is a display node, marked as an output node in ComfyUI, so it sits at the end of a wire purely to show you information. It doesn't modify or apply anything. If you want the trigger words out of the stack as a usable string instead of a pretty report, use Collect Keywords From LoRA Stack - that one hands you the keywords ready to paste into your prompt. This one is for when you want the whole picture including weights and links.

Also worth knowing: none of this needs the LoRAs to be loaded into VRAM. The stack is just data - names and weights - so this is a cheap node to leave in a workflow permanently. The pack's author built the whole stack system around Civitai integration, and while Sage Utils is a small, personal project (the author was candid in a community thread that he mostly uses it himself), the cache design is solid: once populated, it's all local.

Install

ComfyUI Manager → search Sage Utils → install → restart. Manual:

cd ComfyUI/custom_nodes
git clone https://github.com/arcum42/ComfyUI_SageUtils.git
cd ComfyUI_SageUtils && pip install -r requirements.txt

Restart ComfyUI. Only dependency is dynamicprompts. To warm the Civitai cache, run the pack's Model Scan & Report node once on your models folder, or just let this display fill in as you use each LoRA.