Images

Images live in your workspace's assets/ folder and travel with it — upload once, reference as plain Markdown (![alt](assets/...)) forever. Full workflow (upload, resize, align, captions) is in the Writing Guide; Debugging covers broken references and failed generations.

Settings → Images

Providers, looks, and ComfyUI workflows are configured in Settings → Images:

• Providers — OpenAI-compatible endpoints, Stability, FAL, Google Gemini, or local ComfyUI. The editor never cares which one produced the image. Configure credentials, then Test provider before generating.
• ComfyUI Workflows — Import your own API-format workflows into two slots: a text-to-image workflow for Imagine and an image edit workflow (with a LoadImage input) for Imagine again. Each slot accepts an optional seed mapping so every run gets a fresh random seed; without one, the workflow's saved seed is reused verbatim. Your workflows are never modified — margin overlays prompt, reference, and seed onto a per-run copy.
• Looks — Look suffixes appended to your prompt. Shipped looks (Cinematic, Illustration) can be edited, hidden (Restore brings them back), or reset; add your own below the list. None is never deletable and means no suffix.

Generation history

Every run is recorded under History → Images in the assistant sidebar (newest first, refetched on open): thumbnail, prompt, timestamp, seed, provider. Click an entry for full details — input/output images, submitted prompt, paths — plus Copy prompt and Open-folder actions. Delete individual entries with the hover × button; asset files are kept on disk.

Logs live in your workspace at outputs/image_logs/images.json (prompt, submitted text, seed, asset path) — separate from chat ai_logs/.