Stable Diffusion Prompt Extractor reads generation metadata from your AI-generated images and videos and writes it directly into Eagle's annotation field, so you can search, filter, and review your prompts without leaving the library. An easy way to see what prompts and parameters worked and which didn't.

How it works

The plugin watches for new imports and processes them in the background. You can also select any number of existing files and run extraction from the plugin window.

What gets extracted

• The full positive and negative prompts, sampler, scheduler, steps, CFG, seed, model, and image size
• LoRAs used in the generation, with their weights — added both to the annotation and as Eagle tags (e.g. lora: MyStyle)
• Optional cleanup: strip version and epoch information so MyStyle_v2 and MyStyle_v3 collapse into a single tag

Supported sources

• Generators: ComfyUI, AUTOMATIC1111 (including Forge and other forks), CivitAI
• Images: PNG, JPG, JPEG
• Videos: MP4, MKV, WebM, MOV