Start a User Guide for Augmentor

Is there any chance of any documentation coming on Augmentor in the near future? I would really love to use this set of modules to replace what I'm doing with the prompt module that seems to not be actively developed.

I'm purely interested in the integration with ECA, as I'm doing with Prompt now, but I've asked and searched everywhere for some basic guidance which I just can't find.

My use case is simply to send data from one field inside of a prompt to ChatGPT/GPT and store the response in another field. With Augmentor installed there are ECA actions available but I can't figure out how they're intended to work.

I think some basic documentation is really needed for more people to adopt the Augmentor suite of modules.

Thanks in advance!

There is actually the start of a user guide (linked on the project), but it is still a work in progress. I might be able to spend some time on it this week.

Are there particular areas where you're hoping for help? Sounds like maybe setting up augmentors?

Hi @mandclu, awesome!

I'm specifically interested in Augmentor's ECA integration which was added a while ago. I was part of that discussion so I know my use case is possible in the current Augmentor version, I just don't know how.

What I'm doing now with ECA + Prompt modules, is to have automated workflows that trigger on saving or updating new nodes. End users basically provide merely the name of an item, and ECA populates all the relevant fields by sending numerous prompts to OpenAI and storing the results.

General example: After a node is saved, a prompt is sent to AI to write an SEO description based on the body text, and store that in a custom field field_seo_description. Same with the SEO title.
I attached two screenshots of a live example, one showing the model and the ECA action of sending a prompt, and then the prompt that is being sent in the second picture.

Another example is to use a prompt in the spirit of: "Return the calorie count of [node:field_food_item_name)." This is an example but I use multiple prompts like this in a different context to populate new nodes with tons of data instead of end-users having to research that on their own.

In short, when you can send entity data to an AI wrapped in various prompts, and store the output in different fields, you can create super-powerful ECA models and exciting user experiences. You can create a node type "Athlete" and have AI populate all the publically available data from that Athlete just based on a name.

Of course using this, you can also create chat bots, you can evaluate if a comment might be spammy and have ECA unpublish it and notify a moderator, etc. Really the sky is the limit once ECA and AI join forces. It can replace a lot of AI based contributed modules, all you need for almost anything is just this.

Looking forward to any guidance, and more than happy to synthesize my experience into a user guide contribution :)

@AlfTheCat

you can send entity data to an AI wrapped in various prompts, and store the output in different fields, you can create [...] exciting user experiences. You can create a node type "Athlete" and have AI populate all the publically available data from that Athlete just based on a name.

AI interpolator → is very powerful module automatically filling the fields depending on other fields. It skips ECA and does this natively with fields inline. Please checkout the video https://www.youtube.com/watch?v=Sw1Zk__M16c. Possibly it will be quicker and easier to setup.

I promote → ECA and Augmentor in different contexts as well, but this time AI interpolator may be next level simpler for this use-case. Please checkout the video, and let me know how these different ways of doing things compare.

@mindaugasd I watched your video and it looks awesome. I'll enjoy having a go with it over the next few days, I'm sure. I love the batching and cron queueing too, that solves an important issue too. I'm on your newsletter now too :)

Automatically closed - issue fixed for 2 weeks with no activity.