Where communities thrive

  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
Repo info
Needs to point to /auth/twitter/callback in your huginn
I have the above set in the Twitter Developer settings
@comicmuse are you able to provide a screenshot as to what you changed specifically?
Colm Linehan
Nathaniel Stickman

I've been beating my head against this for two days, but I cannot seem to get my Huginn instance to send emails via Mailgun SMTP. Does anyone have any advice?

Whenever I attempt a dry run with my Email Agent, it takes about 30 seconds to process before I get an execution expired error.

Here is my .env with my SMTP configuration:




And here's the command I'm using to run Huginn via Docker: docker run -d -p 80:3000 --restart=always --env-file .env.huginn -v huginn-data:/var/lib/mysql huginn/huginn.

The Doctor
Are you sure that it uses plain auth?
Nathaniel Stickman
Pretty sure, but I'll try login instead and see if that works. I'm also working with my server administrators. It may be that there is a restriction in place to prevent outgoing emails. I didn't think that would also apply to external SMTP services like Mailgun, but maybe it does, since I found that I'm not able to send emails via Mailgun's command line method either. Though I can send emails via curl and Mailgun's API.
Hey, I have an event flow where I detect changes to a webpage. It works, but the only function I'd like to add is to post the changes (and only the changes) to a specific endpoint.
Where I'm blocked is the diff. I've googled a fair bit and I don't see how to have an event just tell me "Yo this is what was different between Event A and Event B".
Can anybody give me a pointer there ?
Hello, I cannot seem to get Huginn to work on Docker and docker-compose and keep the database persistent. Every time I restart, the database is back to the default. I am using the huginn/huginn and mariadb/latest images. Any ideas?
I can connect and login to the database from the command line
@f1yankees I had the same issue. My fix was adding a volume to mysqldata container mapping /var/lib/mysql to some place on the parent hosts disk. Something like this:
    image: mysql:5.7
    command: /bin/true
      - <PATH_TO_WHERE_I_WANT_MYSQLDATA_FILES>:/var/lib/mysql
@sawasy thanks, I was doing that but that did give me another idea as I saw you using the mysql image. For whatever reason I could not get it to work with mariadb but it is now working fine with postgres
Sweet! Glad to hear you got it working. FWIW I am using the bone-stock https://github.com/huginn/huginn/blob/master/docker/single-process/docker-compose.yml with the last two lines stiched on
Hey guys. I'm hoping someone could help me. I'm trying to monitor a URL (only the front-end) for any changes. For example I want to change my wix website's background & that should trigger a change on Huginn. I've created a new agent & have changed the URL to my website (https://willsmalnz.wixsite.com/spykerwolf). However it keeps saying Working: No. Any advice?
@Spykerwolf as far as I know it says not working if it has not spotted anything within the time period that you set it up with in the expected days field.
Sorry, noob question. Does anyone have a simple workaround for the 'mimemagic breaking everything' problem? Fresh install on virtual machine (aws).
Marlin Sööse
Are there any updates on when changes for supporting Heroku-18 will be merged?
Hey folks, I have a question about using xpath in huginn: I'm scrapping a link from a page (xpath: /html/body/div[1]/div[5]/div/div/div/div[1]/ul/li[1]/article/div[1]/a/@href), but I want to scrape all links that have similar xpaths. E.g. I want to scrape not only /html/.../div[1]/ul/li[1]/.../a/@href, but also /html/.../div[1]/ul/li[2]/.../a/@href, and /html/.../div[1]/ul/li[3].../a/@href. Does anyone know of a way to do this?
ah, this was actually an easier solution than I had expected, SelectorGadget helped me find the CSS tags I wanted!
Corey Jack Wilson
Hey, I am running Huginn via Container Station (docker) on QNAP, everytime I login I get "the page you were looking for doesn't exist". Huginn is just using the sqlitedb from a volume on the qnap. I have tried passing in DOMAIN with the qnap ip and the containers port. Is there anything else I am missing that I need?
Quick question: I have Twitter all configured in Huginn, but I want to follow @ and export them to Slack (so I can respond in an appropriate time to people who mention me). How do I do this?
The Doctor
@Jayson-Q What you're asking about would go something like this (at its most simple):
Twitter User Agent following the account(s) in question.
Event Formatting Agent to capture only the parts of the tweet you're interested in.
Slack Agent to post the events to a Slack channel.
Tarun Singh
Hi, I tried deploying to heroku as well as direct docker setup it is not working due to some ruby issue.
Any workaround for this?
The Doctor
Playground Berlin
Hi! Any hint how to capture screenshots?
The Doctor
A visual agent graph debugger... :thinksphere:
hello, something is broken in my huginn server. i created this twitter streamagent, and it still works, like this (in log)
[ActiveJob] Enqueued AgentCheckJob (Job ID: c57a873b-dc22-48c0-87ba-69be01a7aa9e) to DelayedJob(default) with arguments: 10
however when i clone that same exact working agent, then the above ^ message no longer appears in the log for the new agent. and the agent never runs
and there is no failure or other error message in the interface generated. to say what is actually the matter
perhaps it's some rails thing broken somehow?
i can re-download the current latest docker image, (but it may be the same)... perhaps there is some way to rebuild the rail database / schemas etc. too?
it's like i click 'run' on the agent, it creates the job in the job queue. all fine. then when it's the time for the jobs queue to be processed... then it just dissapears from the job queue. and is missing. job is never run, there is no event, no error generated. no log output generated. nothing
how can i debug something like this? how can i switch on any extra level of debugging?
Harpinder Jot Singh
Hey everyone, I am new to huggin and gitter, so forgive me if I ask a question which is not expected to be asked here.
Could someone point me to some docs/readings on the architecture and designs for Huggin.
I need to understand aspects like, how does this scale, how is work coordinated among agents, is there some message-broker pattern, rpc, REST APIs, or what?
The Doctor
There isn't any, I'm afraid, but there really needs to be.
I've only figured out a little bit of it.
How does it scale? If you run a bunch of delayed_job workers, pretty well. I've got Huginn spread across two machines these days.
How is work coordinated? I'm not entirely sure. I know that when jobs are supposed to be executed is stored in the database somewhere. And there is a separate scheduling daemon that queries the database and updates the "when this thing is supposed to run" entries.
I don't know about a message broker; been wondering that myself. No RPC that I know of. There is an internally used REST API (for the Rails front-end) that you can suss out this way:
cd /path/to/huginn
rake routes
Harpinder Jot Singh
Thank you @virtadpt for this.. Things make some sense now.
The Doctor
@SinghCoder_gitlab You're welcome.
Hello, Im trying to integrate Discourse with Yammer so when someone creates a new topic on Discourse we get a notification on a specific Yammer community.
is this possible with huginn?
The Doctor
If Discourse has an API or an RSS feed for topics, you can use Huginn to pick up the new events.
I don't know anything about Yammer, first I've heard of it. But if it has an API of some sort you should be able to use a Post Agent to push events into it.
The Doctor
Has anyone looked at this bug?