comfyui-lora-FindingLora

---

Why I built this

I have over a thousand LoRAs. ComfyUI's stock LoRA Loader makes me scroll a giant dropdown to find any of them, then if I want to stack another I have to drag out a second loader, wire its MODEL in, wire its MODEL out, and remember the trigger words. Every part of that workflow is friction I've felt hundreds of times.

This pack is what I wished the stock loader was: bookmark the LoRAs you actually use; type a few characters to find any LoRA in the folder; store the trigger words once and copy them with a click; chain another loader with a single button that wires itself into the model line for you. Each feature is small in isolation; together they make the most-used node in my workflow feel sharp instead of clumsy.

I haven't seen another LoRA-loader pack do this combination — bookmarks + trigger storage + click-to-copy + fuzzy search + auto-chaining — so here it is.

---

What's new vs the stock LoRA Loader

• No more horrible left/right chevron dropdowns. Both the LoRA picker and the bookmark picker open a proper modal on click. Default view is alphabetical with the current selection highlighted and scrolled into view, so a thousand-LoRA list lands you near where you were. Start typing to fuzzy-search — same picker, no separate search button.

• 📖 Bookmark button. One click bookmarks the currently-selected LoRA. The button label flips to 📕 Remove bookmark when the active LoRA is bookmarked.

• 📚 Bookmarks bar. A second clickable bar above the LoRA picker shows the active bookmark and opens a picker modal listing every bookmarked LoRA. Picking one sets the LoRA picker — the bookmark and the active LoRA always agree.

• ✏️ Trigger word storage. When you bookmark, you're optionally prompted for a trigger word or phrase. The trigger displays on a read-only line when a bookmarked LoRA is selected, and gets emitted as a STRING output you can wire into your prompt-encoding chain.

• 📋 Click-to-copy trigger. Click the displayed trigger row to copy the phrase straight to your clipboard — paste it directly into a CLIPTextEncode. Shows a quick toast confirming the copy.

• 🔗 Chain another LoRA Loader. A button at the bottom spawns a fresh copy of the node beside the current one and splices it into the model chain. Any downstream MODEL connections get re-routed through the new node — (upstream) → this node → new node → (former downstream). Stack as many LoRAs as you want without manual rewiring.

• ➕ LoRA stack — up to 10 more LoRAs in the same node. Below the strength slider, a green ➕ Add another LoRA row opens the same fuzzy-search picker and adds a stacked LoRA with its own strength. Only the rows in use are shown — one add row at first, growing to a maximum of 10 stacked LoRAs (plu