Huginn: System for building agents that perform automated tasks online

Do you have a writeup, or know of writeups by others, that go into detail of your use? Always a bit hard to imagine given universal tools like this.

I have zero Ruby knowledge, how useful will it be for me? Can it all be done with drag and drop?

> beaten only by crows that day

Very funny, since "Huginn" and "Mininn" were a mythological pair of ravens [0].

[0]: https://en.wikipedia.org/wiki/Huginn_and_Muninn

I'm CEO of robusta.dev (an automations tool for Kubernetes) and this has been our experience too.

People don't want the ability to write their own automations - they want something that just works out of the box.

Very early on we realized this based on feedback and pivoted to focusing on out of the box alert insights (powered by automations we wrote) and not write your own automations.

I don't know what the equivalent is for the web.

> The more generic something is, the less useful it is.

Funny how different people reach different conclusions :)

In my experience, tools and services that don't generalize enough (or allowing it rather), are less useful in the long-term and just solves something short-term. But it's easier to get traction if you solve something short-term, as it's often easier to market something that is more focused.

Wow, Temporal is an interesting find.

I thought of creating a similar tool just for marketing automation. I built a landing page [0], but haven't had the time to talk to potential customers.

[0] https://h11g.com/

It kind of reminds me of Yahoo Pipes

I agree, I've never quite been able to figure out anything useful over alternatives. Like the usual "send me an SMS when the weather will be rainy" example - I use a highly customizable weather app that gives me more power to send me stuff like that without having to fiddle with a new system.

This system's integration with Mechanical Turk could be interesting, but I'm too dumb right now to figure out some cool things I could do with it.

I have “frenemies” who are semi popular. I want to keep alerts on their LinkedIn and Twitter. Similarly startups I want to know if they’ve stagnated or gotten funding. I would like to get earnings calls emails for some companies I like. I want to get a ping if the forex rate changes for some combos, or if gold price drops by some percent. I and family have many bank accounts and it’s a real thing that you need to keep an eye on the finances of your banks here (one goes down every few years). I will like to have news alerts on all the small towns my family is in so I know if something big happened anywhere.

I think it’s definitely a mindset to think of everything you do or read every day and think if you can offload it to someone or something else. Once you switch that knob on you notice so much every day.

I setup huginn about two years ago primarily just because it looked interesting. I then sat there for a while wondering what to do with it. I setup the basic things like a morning weather alert, and daily digest email and basically ended up with the same feeling. I decided to leave it for a couple weeks or so and just see if I came up with anything else more useful. That was definitely the right way for me to get going with Huginn. It's ultimately a bit of a multi tool, if you keep it around you're bound to reach for it although you may not need it the second you buy it.

Since having it setup I've used it to:

* Monitor covid travel restriction changes on the UK gov site and email me when updates are made.

* Scrape the site of a housing development where I wanted to view a house early as a potential purchase to know first when they opened viewings.

* Look for a price drop on some computer monitors on Amazon as the price was quite high when I found them.

* Daily discord report on the price of flights on a particular date when I and a group of friends were thinking about a holiday together.

* Alert when an exchange rate dropped or increased by a certain amount to help figure out when to exchange money for the same trip.

I find most of my uses include some kind of web scraping, although not always, and are usually things I want to run frequently and have run reliably. Most don't live more than a few months but it's a great tool to have knocking around as the effort to spin up an automation is much lower compared to writing a custom script each time.

I actually have n8n.io self-hosted as well because it has some nice pre-bundled integrations with various services like Spotify which huginn doesn't. That said, I've not really enjoyed using it, I find myself hitting F5 a lot because the UI breaks and won't recover and while it's more visual it's actually much harder to view the logs of a workflow. I tend to default to huginn unless n8n has a particular integration which will make the task faster to setup. Once setup though, I've not had any problems with it.

I've used it exclusively for years, and it's produced some of the best results.

Looks like a web version of RPA.

N8N is awesome for some tasks, but infuriating for others.

The time I've spent trying to filter seen data from a source is just mind-boggling. It would've been a lot faster to do the whole thing myself.

And I'm not alone in this. Pretty much every task that updates on a timer needs to exclude things it has already seen - and this is a herculean cryptic task in N8N.

I definitely think there needs to be some kind of sea change for this kind of tool to really have an impact. I fret that there's not enough middle-brow interest in tech & computers for even the best designed, most capable user-scripter (like Huginn or Node-RED) to become truly high impact. Developers tend to find a lot of value add versus deploying whatever industrial ops practices they know. So whose left to support & nurture these interesting vanguards, these more comprehensively-built & user-focused computing environments?

And reciprocally getting started with a project like this is somewhat daunting for the newcomer, while at the same time they're bombarded by far more options & places to get started in computing than they can handle. I think most folk kind of swear off involvement, don't have a way to casually check out a couple random computing attempts easily, dabble & sample. Computing rebuffs interest, alas.

I still believe in change. The dis-interest in computing is obvious & clear to me today, because most computing is back on shitty mainframes (cloud) in places we can't see or poke at (apps, cloud). Trying to better merge the developers back into the folk would help a lot. But real wins are needed broadly to tell users they're wanted on the other side, to make exploration fun and cool again.

When I looked into it some time ago, what tripped me up was that it seemed that you needed to deploy it using Docker. Otherwise, you'd have quite an involving set up process: install Ruby, Rails, then a database. Then you'd probably want to figure out a set up to enable https + SSL. I had it running locally and was quite impressed by it, but wasn't ready to go the full Dockerized route on a VPS at that time. Maybe now I would.. just need to find the time.

Personally, I think a project like Huginn would have more success if the code was in GoLang (easier to deploy) or Python (more popular language = people more willing to set it up) and used SqlLite for the backend. And if it had SSL/https baked in, out of the box. For comparison purposes, I distinctly remember how much I appreciated that Nextcloud had it baked in. So I ran NC very happily for quite a while and skipped out on Huginn.

I use it daily at my law firm. It monitors online court dockets for many of my lawsuits and sends me an email when there is a new entry/order/filing. Incredibly useful.

yes, i love the thing - i've been thinking of setting up klaxon from themarshallproject for the web diff/watching stuff i have - but still haven't looked in to it deeply (and not sure i like their ready-to-go deployment without flexibility)

Examples of what i do with huginn:

1) twitter watching for peaks in activity in keywords around my home country (this one is hard, football interferes with it)

2) pulling in data from an API and storing it (where i jury rig a grafana dashboard to read from a view over the huginn data)

3) stuff that sends me push notifications on certain events (a bit of a rube goldberg machine, but: homeassistant zigbee device action is to hit a webhook on huginn, that webhook sends a push notification through huginn + alexa notifier event)

4) I have a zigbee panic button outside the house (again, attached to home assistant) in case someone is locked out without a phone and without keys that will then hit a webhook on huginn that will call the other adults of the house with a message

5) keep an eye on the job pages of a few companies i was interested in, filter out for roles with keywords in the title, and email me the details.

6) I'm waiting for certain products to become available, i have huginn watching those pages and sends me an email if it becomes available (if the manufacturer doesn't have a 'let me know when available' function) - this is unfortunately quite custom for each site, but do have a common chain afterwards

7) there's an english language news site that doesn't have rss or any form of notification, so i scrape their headline ticker and send myself push notifications

It's not necessarily as flexible as i'd like - but it gets enough of the job done often enough, for anything else i can feed in webhooks from custom scripts, etc.

If you get used to thinking in huginn, you will also be able to remove a lot of duplicated work and create common chains/funnels for events, that way you'd only need to add in one new agent at the top of a chain and you've already done the rest of the work.

edit: formatting