[Feature Request]: Allow attaching a reusable reference image set to image-generating models

[MantridHands]

Summary

It would be very useful to have a way to define a persistent, reusable set of reference images in the Agent configuration, so that image‑generating models (OpenAI Image Tools, Gemini Image Tools, etc.) can always use them as visual context, without manually re‑attaching images on each message.

---

Current Behavior

Right now, images can be:

• Uploaded per message and optionally used as context / for editing.
• Referenced via image_ids when available in the current context window.

However:

• There is no way to attach a fixed set of reference images at the Agent level (similar to file context/RAG), so that every conversation with that Agent can consistently use them as visual context.
• For workflows like “brand style guide”, “clinic interior references”, “character design sheet”, etc., users have to repeatedly upload or re‑attach the same images in individual chats.

---

Desired Behavior

For Agents that use image tools (OpenAI Image Tools, Gemini Image Generation, etc.):

• Add an option in the Agent config to upload or select a reference image set (one or multiple images).
• These reference images would be:
  • Stored once and associated with the Agent.
  • Automatically made available as image_ids (or equivalent) for the image tools in that Agent’s conversations.
  • Optionally visible in the UI as “Agent reference images” so the user can see/edit the set.

This would allow, for example:

• A “Brand Image Agent” that always uses the clinic’s logo, color palette, and interior photos as visual context.
• A “Character Designer” Agent that always has the base character sheet in context.

---

Scope / Components

• Agents configuration UI (Admin panel & user Agents).
• Image tools integration (OpenAI Image Tools, Gemini Image Tools, etc.), passing Agent‑level reference image_ids along with per‑message images where appropriate.