Today we are releasing AI Action, a new plugin for Eagle.
AI Action brings a new way to organize assets. You hand your own organizing rules to AI and let it apply them in batches. There is no more naming, describing, sorting and tagging one image at a time. You create Actions, choose which steps to run, which model to use and which custom instructions to follow, and AI organizes your assets your way.
With AI Action you can:
- Rename files with AI: rename them based on what the image shows
- Describe files with AI: analyze an image and write the result into the description field
- Sort files with AI: place them into your existing folders automatically
- Tag files with AI: add tags following the logic you set
These can be combined freely. Create an Action that only renames, one that only describes, or chain naming, describing, sorting and tagging into a complete flow. To us, the biggest point of AI Action is that your way of organizing becomes a workflow you can run again and again.
Why did we build AI Action?
As assets pile up, organizing usually gets stuck in two places.
First, organizing takes time. You may know these situations: you import a batch of images and never get around to naming them; you know an image is worth keeping but never add a description; you want to sort and tag, but once there are many files it is hard to finish. The images keep accumulating and the library gets harder to search.
Second, everyone organizes differently. Some people care about file names, others rely on descriptions; some use folders, others mostly use tags. Some care about photographic language, others about brand tone or how a piece will be used in a proposal. We have always felt that organizing assets should not have a single fixed answer.
AI Action addresses both problems. You hand your way of organizing to AI to run in batches: you define the rules, AI does the work. Organizing can move from a personal habit to a method you can reuse.
In "Rethinking how we organize images in the age of AI", we broke down the thinking behind AI Action's design, if you would like to read more.
What is AI Action?
AI Action is an AI workflow tool you can combine freely. You create different Actions and specify what each should do. An Action can handle a single task, or chain several steps into a complete flow.
For example, you can create an Action that only renames images, or one that only writes descriptions; one that sorts images using your existing folder logic, or one that tags by style, color and use. Chaining naming, describing, sorting and tagging together works too.
So AI Action is not a single feature button but an organizing tool you design yourself. What matters is how you want AI to work for you.
Rename files with AI: clearer, more consistent names
When images enter a library, their original file names often carry little meaning. They may be download gibberish, camera numbers, or a name that made sense at the time and is forgotten a few days later. As files pile up, the file name becomes the layer of asset management that is easiest to ignore and does the most damage to efficiency.
AI Action can generate more recognizable names from the image content. You can have it describe the subject, bring out the tone or style, unify the language, add a prefix, respect a length limit, or apply your own naming rules directly. Some people want Traditional Chinese, others lowercase English; some want every file name to start with REF_, others want names written from a photographer's view with composition details, or kept within a fixed length. All of this can be set through settings and custom instructions.
As your library grows, a consistent and recognizable naming scheme is easier to maintain and search than scattered original file names.
Describe files with AI: add text that is actually useful
If AI renaming builds a more intuitive outer layer of recognition, AI description adds fuller text information, giving each image another layer of content that can be understood and searched.
You can ask AI to describe the subject and scene, analyze the main components and layout of a UI screen, analyze the composition, light and mood of a photograph, record brand tone and suitable proposal situations, extract text from the image, or focus only on the angle you care about. This is also the most extensible capability in AI Action. Often the real difficulty with an image is whether you can find it later, remember it, and bring it back with words.
Once images have clearer descriptions, finding them again is noticeably faster, whether you look yourself or combine it with Eagle's search. Put simply, AI description builds semantic information for an image so it is easier to find and use again.
Sort files with AI: follow your folder logic automatically
AI Action also supports automatic sorting. AI can place assets into existing folders based on their content, saving a lot of manual dragging and case-by-case judgment.
This works especially well when you already have a clear folder structure. The clearer the folder names, descriptions and boundaries, the more stable AI's results tend to be. For example, if your folders already separate UI component screenshots, brand visual references, product photography, commercial portraits and typography cases, AI sorting can save you a great deal of manual filing.
AI sorting suits the case where you already have a clear folder structure and filing by hand simply takes too long. The more mature the structure, the better the result.
Tag files with AI: hand over the task that is easiest to put off
Tagging is valuable, but easy to postpone. Not because it is hard, but because it takes time. You have to look at the image, think of words, decide whether it deserves a tag and choose the category. One or two images are fine; with many it is easy to keep pushing it back.
AI Action can process this in batches. Following your own tag system, AI can analyze color, style, subject, mood, use, composition and any other dimension you care about. The clearer your tag groups, naming and description rules, the more useful the tags.
The value is that the organizing work you know matters but never finished can finally run in batches. The larger your collection, the bigger the gap between doing it by hand and in batches.
If you want to use AI tagging or AI folders, we strongly recommend installing the AI Search plugin first.
AI Search builds an image-similarity index for your library. When AI Action detects that data, it can use visual math to greatly improve the accuracy of sorting and tagging. That means AI does not only analyze image content objectively; it can learn your sorting habits and produce results closer to your own logic.
Custom instructions: where AI Action gets powerful
If we had to name AI Action's core capability, it would be custom instructions: you define how AI should understand your images.
When analyzing images, different roles need very different things:
- UI designers care about interface type, components and layout
- Photographers care about composition, light, tone and how the subject relates to the frame
- Brand designers care about mood, tone and how a piece will be used in a proposal
- E-commerce teams care about product type, shooting angle and categorization
If everyone had to use one generic way of analyzing, the result would stay "roughly right" and be hard to drop into your own workflow.
Custom instructions let everyone define AI's direction of analysis in their own words and logic. You are telling AI how these images should be seen, described and organized to fit the way you work.
Your way of organizing becomes repeatable
What we find most interesting about AI Action is that it turns organizing habits that used to live only in your head into something that can run again and again.
You may already have your own method: how files are named, how descriptions are written, which images go into which folder, which tags apply. Until now you could only do this yourself, one image at a time. Now you can write those rules into an Action and hand it to AI.
Your method becomes a system that runs. Organizing is no longer a one-off effort; it can build up over time.
Who AI Action is especially good for
AI Action is especially good for:
- People whose collection is too large to organize by hand
- People who already have fixed habits for naming, describing, sorting and tagging
- People who want to fix their own organizing logic and apply it repeatedly
- People who want to fill in text information first so later searches work better
- People who are not satisfied with generic AI analysis and want AI to work their way
It delivers the most when you already know how you want to organize, are willing to spend a little time writing custom instructions, and already have some folder or tag structure. For sorting and tagging especially, the clearer your existing structure, the more stable the results.
AI Action is not meant to replace your organizing system. It makes the system you already have run faster.
Used together with AI Search, the result is more complete
If AI Search solves how to find things, AI Action solves how to make assets easier to find.
The relationship is direct: use AI Action to give images fuller names, descriptions, folders and tags, then use AI Search to bring them back. When the images carry more information, hit rates rise whether you search in natural language, in text fields, or by image.
Organizing and searching are two sides of one thing. AI Action supplies the data; AI Search turns it into the ability to find.
Start letting AI organize for you today
AI Action is released today. You no longer have to choose between organizing everything by hand and handing everything to a generic AI. You can keep your own logic and let AI run the rules in batches.
If you have built up a lot of assets in Eagle and have long known organizing matters but never had time to do it well, AI Action is worth a try. It finishes, in batches, the things you already know how to organize.
To get started quickly, read the AI Action quick start guide, which walks you through your first Action from scratch.
To understand the logic and efficiency tips in depth, read the AI Action best practices guide. It starts from four core ideas, breaks down how the features are used, and offers custom instruction examples and model selection advice to help you build your own AI workflow.