For organizations that only use this internally (not an MSP), it would be great if SSO allowed for group access based on group within Entra. So for example, one app in Entra. But if user in in Entra Group 1, it gives them access to companies A and B in Hudu and groups A, B, and C. Then if a user is in Entra Groups 1 and 2, it gives them access to companies A, B, and C and groups A, B, C, and D.

A critical part of the SSO should be that it works with the client's M365 Tenant. This is huge with HaloPSA - no new accounts at all, the client's own tenant and creds work.

Would love to see this

Hi Jarrod! Yes, we have a fairly robust API. You can find our Swagger documentation page by going to the Admin >> API >> View API documentation. It's not 100% feature parity with the UI, but it's close – pretty much anything you'd want to do with outside integrations should be possible (or if it's not, feel free to create a feature request at feedback.hudu.com).

We've built out the Automation Library inside the community (this portal) so that customers can share their examples/what they're doing with it.

Out of curiosity, what integrations are you looking for? You can also request those on our feedback site.

Tons of these different mermaid visualizers popping up, but I also +1 on mermaid.live specifically because they only store the diagram in the URL 😛

This event (https://thehackernews.com/2025/11/years-of-jsonformatter-and-codebeautify.html

) has been in the back of my mind recently with a lot of our guys finding different tools and bringing them up for internal review

Thank you for the share!

This is awesome. I would have never taken the time to do this myself.

Thanks for sharing!

Update, I did hear from the Dev team and we are aware of some performance issues that are slated to be improved in 2.44.

If you haven’t already, it still may be worth filing a support ticket.

Hey Scott!

I assume you’re self-hosted?

I haven’t heard of any issues, but it’s always possible. I’d put in a ticket with support here or [email protected] — especially if you’re cloud hosted, there could be a resource issue with your instance.

Mason, thank you for such a thoughtful reply. I'm sure any of these 4 options would solve the issue. Before going down one of these roads, is it safe to say that I cannot run my migration test without doing at least one of these options? This isn't a production environment so I'm looking for a shortcut to test the import options. Thank you again.

You can indeed do a self-signed cert if need be, but I wouldn't recommend using the placeholder certificate personally, but there are some easy alternatives.

Option 1- Create an NGrok tunnel

This will upgrade your http connection. You can do this by installing ngrok (available as a snap package too, which makes things easy), and create a tunnel with

ngrok http 80

it will then give you a FQDN that you can access 10.0.0.10 over TLS, but the free version is subject to limitations (below). then you just set migration config to migrate to the FQDN from ngrok. It's pretty cheap as a one-time expense if you want to go the easy route

Option 2- LetsEncrypt / Certbot method

This method just requires you to have control over a domain name and admin access to your router.

A lot of the times, it's easier to get a certificate, however-
if you are on a home network, you can forward (external-interface only) ports 443 and 80 to 10.0.0.10 443/80 respectively on your router.?

then, you will want to add an A record on that domain (or subdomain) that points to your external IP address (with the hostname of device at 10.0.0.10. Then, you can sign in on 10.0.0.10, request a certificate from LetsEncrypt, and that certificate will be legit no that certificate will be trusted by browsers and operating systems, provided clients connect using the hostname the certificate was issued for.

After which, you'll want to un-forward the ports there and probably set your migration to point to the fully qualified domain name that you certified. (somecomputer). lastly, you'd have to (depending your local DNS setup) add a hosts record temporarily that tells your computer that the fully-qualified domain name you certified resides at local address. (this is because of X509 host validation)

to be succinct:
your local machine says that 10.0.0.10 is where FQDN points to
your DNS says that external ip address of your net is where FQDN points to
your router forwards ports 443/80 (just during time of certification) to 10.0.0.10
certificate gets installed on 10.0.0.10 and is added to nginx config.

Alternatively, if you can provide domain control verification to a different issue via TXT record, that makes it so you dont have to forward any ports

Option 3- Self-Signed

If you have an on-premises certificate authority, especially one that is in your current trust-store, that is going to be pretty easy.

Otherwise, acting as a temporary certificate authority and signing one should work too, so long as you add the certificate issuer (yourself to your trust store).
Below is a (rather-large) oneliner to do just that. In this example, we're certifying test.local (which you'll have to set your dns to point to the LAN address for this.

(be sure to adjust mylocaldnsforwebservice to be the actual hostname of the device @ 10.0.0.10)

$ca=New-SelfSignedCertificate -Type Custom -KeyUsage CertSign,CRLSign,DigitalSignature -Subject "CN=Local Test CA" -KeyExportPolicy Exportable -KeyLength 4096 -HashAlgorithm SHA256 -CertStoreLocation Cert:\CurrentUser\My -TextExtension @("2.5.29.19={critical}{text}CA=true"); $srv=New-SelfSignedCertificate -DnsName "mylocaldnsforwebservice","test.local" -Signer $ca -KeyExportPolicy Exportable -KeyLength 2048 -HashAlgorithm SHA256 -CertStoreLocation Cert:\CurrentUser\My; Export-PfxCertificate -Cert $srv -FilePath .\server.pfx -Password (ConvertTo-SecureString "password" -AsPlainText -Force) | Out-Null; Export-Certificate -Cert $srv -FilePath .\server.cer | Out-Null; Export-Certificate -Cert $ca -FilePath .\ca.cer | Out-Null; Import-Certificate -FilePath .\ca.cer -CertStoreLocation Cert:\CurrentUser\Root

If you do self-sign, acting as your own CA, it's important that you remember to remove the temporary CA certificate from any trust stores where you imported it once the migration is complete.

Get-ChildItem Cert:\CurrentUser\Root | Where-Object Subject -eq "CN=Local Test CA" | Remove-Item; Get-ChildItem Cert:\CurrentUser\Root | Where-Object Subject -eq "CN=mylocaldnsforwebservice" | Remove-Item;

Option 4- if it gives you trouble, you can get an inexpensive droplet from DigitalOcean for quite cheap (get cert from certbot here, requires domain name)

In all, there are a few ways to do this in your home lab or on office machine. personally, I'd just get an SSL certificate from Letsencrypt on my LAN or on a Digitalocean droplet.

Hi Elliot Kamper! This was an oversight due to the generic "can't delete" policy for non-admins that we have in place, that we unfortunately did not catch.

The team is aware and has resolved this; we have it scheduled to release in the 2.43.2 update.

I have two child roles under NMS_Other, one for IP Cameras and the other for Sensors. I would like to be able to sync these to Hudu

Kris,

Thanks for sharing that!

The behavior you’re seeing is expected because these devices are assigned the NMS_OTHER role. The Ninja integration currently skips that role intentionally, since Ninja uses it to denote NMS IP addresses.

That’s why the devices do not import when mapped as NMS_OTHER, while mapping them to something like NMS_Firewall allows them to sync successfully.

I hope this helps. Let us know if there are any questions.

I'd really like this too....as we have our own ollama server inhouse. Initially I just figured I'd proxy it through Open WebUI but you can't change the URL for OpenAI in Hudini.

We're currently using local LLM to clean up our KB article formatting. So far so good but rate limiting can catch you out a bit.

Hi Elliot Kamper!

Thanks for this discussion. We’ll look at adding support for bringing phone numbers into Hudu via the Google Workspace integration.

Appreciate you bringing this up and starting the conversation around how others are handling staff contact information in Hudu.

Best,

Figured I would give a quick update on what we have going now. Lars sent me down a bit of a rabbit hole (Appreciated!) and gave me the kick I needed to completely review and redo our current setup.

I ended up finding out during the setup review that while we were doing nightly VM snapshots on Azure, the old DB backup script had not been working in over two years. That was a bit of a wake-up call to completely audit the entire setup from start to end.

I've now got this backup setup tested and working flawlessly. Still not perfect and will be tweaking things a bit more, but it should be good for now.

What I did with a little (ok, a lot more than I care to admit, my linux is a bit rusty) help from ai:

• Cleaned up and improved our DB backup to our Wasabi backup bucket. It now runs every 4 hours and notifies us daily via webhook to Teams channel on backup status.

• Full backup and sync of our Wasabi bucket that includes both DB backup and all uploaded files from Hudu to our local Synology NAS every 4 hours.

• New on-prem VM hosted on Hyper-V that works standalone in case of downtime at Azure. We just make a DNS change and then we are hitting the internal Hudu instance. Longest data loss window is 4 hours, and we have full HTTPS/access locally.

• The DR backup download and restore to the VM happens every 8 hours.

We also upgraded our Azure VM a bit as it was long overdue for some cleanup and changes.

I may further tweak things a bit as of right now, the DR still relies on our Wasabi instance for any uploaded files/pics storage and want to remove that failure point from the DR VM. We are pulling those files down anyway as part of the daily backup sync to our local NAS, so it's just a matter of changing how we point the DR VM to that storage location. Since upload files are not as critical, I was not as concerned, but it does need to be addressed.

We have an instance running in a VPS provided by Ionos. Super cheap and downtime has been nonexistent other than during Hudu updates. Of course we have backups running to S3 as well just in case.

Hey mw-fisit , I see our Support Team got you taken care of.
I updated the post to let everyone know that this is resolved with Hudu Version 2.41.2.
Thanks for reporting it to us!

Hey mw-fisit,

I have picked up your ticket, and we are urgently investigating the issue and working on a fix.

This is our top priority. As I receive more information, I will provide updates directly in our email chain.

If you have any questions, please do not hesitate to reply here or in the email chain.

We will continue investigating and keep you up to date.

Thank you,

Hi Adam Hough!

Sorry to hear you're not receiving alert email notifications.

It sounds like everything is set up correctly. To rule out a couple of things, is SMTP configured and working correctly? Also, is the alert set to stop after reaching the trigger date?

Hello Adam Hough. Sounds like a bug, but let me get our support team to help you answer this fully.

Chrissy Meyer, can you or someone on your team review this and then update our support articles if needed?

That's a great question.
There's definitely a few areas of consideration-

For group checklists with tasks that might need to get reassigned or re-delegated, I'd err towards Hudu. Autotask is great, but the inability to assign different tasks to individual users makes it a bit too 'rigid', in my personal opinion.

The really nice thing about Autotask is that it can be directly linked to contracts, billing, tickets. For items that are deeply reliant or synced-in with those business aspects, it might make sense to continue using Autotask checklists there.

if there are any mission-critical processes or tasks that aren't tied to to procurement, contracts, or tickets, I think Hudu Processes would be the way to go. Largely for the fact that you can drill into activity logs when something about a process or task... smells funny. While Autotask does offer broader support for auditing ticket items there are workarounds to effectively log individual task activity, it's somewhat weak in that area in my opinion.

TL,dr;
for CRM and service-desk stuff where billing, ticketing, and timekeeping is a major concern, autotask makes sense to use, in my opinion.

For internal processes or client processes that are technical, require audit logs, might have tasks reassigned, and tie into devices or 'things', I think Hudu processes make more sense. These are great for ensuring consistency when initializing client devices or onboarding a new employee for a client (who will need to be documented), as example.

Like anything, there's always a balance.
Just my 2 cents!

Hi team

Do we have any update on this issue?
I just removed the extension and added it again, but seems the issue still continues unfortunately?

Thanks,

Andrew

Hi Jordan

Thanks for the update, I've tried as per Michael's suggestion to remove and add the extension.
Unfortunately the same issues occur.

In incognito mode I don't even get a prompt to enter the URL to sign in again, so it doesn't work in any way, shape or form when in incognito mode.

I also just noticed this morning. Even when using in normal browser mode, then the extension needs to be re-linked every time the browser is closed and opened again. As if it doesn't hold any settings being applied to it. This is only in Vivaldi browser, all other ones seem to work fine.

Hope that helps and adds some more context

+

Hi Wade. To create a GPT via the API, you'll want to make sure you're using company_id = null. Something like below should work for you:

{
  "name": "Jordans API Process",
  "description": "Standard onboarding checklist",
  "company_id": null
}

Please let us know if it continues to give you issues, however.

Hey Jordan Hart , so, the removal of the tagged passwords was intentional design? I want to know so I can relay that to my team for what to expect.

Thanks!

I agree, this is terrible design for using tags and I assume the upcoming labels feature will use the same method. Anything that can be filtered should also be able to be added as a column. Clicking on the tag column should then show the list of tags. All filtering should be done in the header row and not is a separate side panel.

Another option would be to allow filtering from a shortcut key like the Company and Password Searching is. Like pressing F would bring up filtering and then you can filter by tags there. I know I'm not alone with preferring a product that has powerful keyboard shortcuts.

I also am unsure the benefit of filtering multiple tags, not sure I've ever had a need to do that. What passwords layout actually needs is a proper Location field, since over all multi location company support is lacking.

Hi Robert, thank you so much for the kind words! It was a big release and we are very excited about it. Agree that archiving photos is a huge help!

We have some big things in the works for our new labels feature and are actually planning to do a build a feature post here in Community soon.

Labels will be similar to our current password tag feature but much more powerful. Admins will be able to create custom color coded labels scoped to specific companies and record types (not just passwords). You'll be able to filter company and global record tables by labels and filter search results by labels.

Would love to hear what additional capabilities you'd like to see for them!

Hey Issac! displaying images on the table view is not currently possible.
I looked around in our feature request bored and I don't think this is something that's been asked before, so I'd suggest putting in a feature request with us! You can submit one here: feedback.hudu.com

Hey Terry / Steve G !
Sorry for the delayed response here. That is correct, there isn't currently a way to fully delete a user (because they're referenced in activity logs, configuration logs, etc.).

I am asking the team if there's a way for us to "forcefully" do it from the backend. Data removal requests is a valid concern.
I will pin a follow up comment when I get a definitive answer.

Hopefully Hudu folks chime in on this - I have employees that I wouldn't like to see in there either, but its not causing any harm I suppose

Right now I created a PS script that gets the IPs DHCP status from the IPAM, and then updates a Network Summary magic dash that has VLAN info, DHCP scopes, wireless ssids and network services for quick checks.

We have an asset for it. We came from IT Glue so we had network assets in there, and moved them to Hudu and just continue to use it. We have 2 actually and sporadic use of both. If the network is complex we just use an excel doc. But like above, we have the asset which we use the following:

We will be merging them to a single asset here soon as we revamp our docs. We also have a lot of Unifi and the integration there will insert a card with the data for each network... That ignores the IPAM, and needs an asset layout to attach.

Hey Babak , I would think that has to do with wherever it's being embedded from.
I just tested with a YouTube video and I did have the player controls.
I'd recommend writing into Support (with screenshots) and they should be able to help you troubleshoot!

You potentially could set up a Minio server/service or appservice in Azure and have it write to a BLOB storage or Azure Data Service mount.

It sounds complicated, but azure makes it somewhat painless, they may even have a prebuilt Minio image to select from, though in that case it would likely be the commercial version instead of the free one.

This setup wouldn't be ultra ideal, since you'd effectively be paying for storage twice (although backup redundancy is always good).

It could certainly be done, however

Hey Gregory!

I'd recommend upvoting our feature request for this here: https://feedback.hudu.com/feature-requests/p/ability-to-backup-export-to-azure

Just to bring more visibility.

Nick Ellermann, apologies for the very delayed response here, but I do see that Reddit pointed you to this as well. Unfortunately we don't have any specific videos after this, but if you have any specific questions, we may be able to answer those for you.

Intro Video Series:

https://community.hudu.com/academy-3gsikl8p/post/intro-to-n8n-rJkIihLv8z9xJVm

A few n8n Workflows:

https://community.hudu.com/script-library-awpwerdu?feed.view=n8n

Any online videos or demonstrations we can see of this in action? We have slowly been building out N8N workflows, very few with Ai ties but trying to do more. I finally figured out how to load this into our environment, but clueless thus far how to make use of it.

Hi! Tags are automatically generated based on the most frequently used words in the article. We maintain a list of common words to skip (i.e with, from, about, into, etc.) and skip anything 3 or fewer letters.

Hi Jordan,

Thanks for your prompt and detailed responses!

I have another question/more details regarding the flow of moving articles in Halo and expecting the Hudu updates to sync.

Since the Hudu integration creates all FaqLists and Articles within the FaqList 'Hudu Company Kbs', I was attempting to move articles outside of that FaqList to another high level list. Could this be the cause of why the integration can no longer access the article when its moved outside of the Hudu list?

I don't believe we can set up api access per FaqList in Halo - the agent running the integration has admin permissions, and the integration has read, edit for kbs.

Hi Terry — great questions. I'll try to provide insight as best I can.

TL;DR:
Halo enforces a unique thirdparty_id for each synced Hudu article, so it will never create a duplicate from the same source. If an article is moved, hidden, or placed somewhere outside the API key’s permission scope, Halo still considers it the same record and will attempt to update it rather than recreate it — which can result in errors if the API cannot access it. Deleting and recreating the integration resets those associations, which is why all articles get created again. Halo is also rolling out restrictions on editing synced articles, which should help prevent these sync-breaking changes going forward.

What you’re seeing aligns with how the integration currently behaves rather than a misconfiguration. Both behaviors you’re seeing are related to how Halo tracks synced articles using a thirdparty_id.

When we sync an article from Hudu into Halo, we send a unique thirdparty_id tied to that source article. Halo enforces uniqueness on that field, meaning it will not allow multiple articles to be created from the same source. If a sync runs again, Halo will always attempt to apply updates to the existing article associated with that thirdparty_id.

1. Moving the article to a different FAQ list in Halo

If the article is moved inside Halo, the thirdparty_id remains attached to that record. From Halo’s perspective, that article still exists and is still linked to the original Hudu source.

On subsequent syncs, Halo will attempt to update that existing article rather than create a new one. However, if the article has been moved to a location that the API key does not have permission to access, the integration may not be able to properly reconcile it. In that case, errors can occur and a new article will not be created because Halo blocks reuse of the same thirdparty_id.

Under normal permission scope, I would expect the integration to move the article back to its original synced FAQ list. If that is not happening, it is likely related to visibility or API scope restrictions; something I can have the team dive a bit deeper into.

2. Setting the article to inactive / hidden

This follows a similar pattern.

When an article is marked inactive or hidden, it may no longer be retrievable via the API key’s scope. However, the record still exists in Halo and retains its thirdparty_id.

Because that ID is still registered:

• Halo will not allow a new article to be created

• Sync attempts will try to update the existing record

• Errors can occur if the API key cannot access that article due to its state

Since visibility changes are not communicated back through the integration layer, Hudu has no signal that the article should be recreated.

3. Why deleting and recreating the integration works

Removing the integration effectively resets the association between Hudu articles and Halo records. When reconnected, all articles are treated as new and synced again, which results in fresh copies being created.

Halo is currently rolling out a feature to restrict editing of synced articles (it recently entered beta). Once fully live, we’ll align the integration behavior to prevent structural changes that can break sync alignment.

Congratulations to the winners of the MSPGeekCon drawing! We are looking forward to connecting with you all in May 😃

We've just enabled this and noticed similar issues, though I think this is going to be a challenge without some changes in Halo. I'm happy to go through all FAQ lists that were created and set them to FAQ Groups where needed, however in the instance where you have articles in a folder as well as subfolders in the same folder you lose access to the KB articles in the folder if you set it to a FAQ Group, but if you leave it as a FAQ List only then you lose the subfolder.

Hudu could work around this by picking up on this and creating a special "root" FAQ list inside the FAQ Group for those documents, though I will admit that's not the cleanest solution.

Hello Steve G!

Thank you for the feedback on the HaloPSA KB Integration!

For the “Use this FAQ List for grouping other FAQ Lists” setting, this is something we’re aware of and have an open report to address. The goal is to improve how this is handled so it doesn’t require manual adjustment during sync.

Regarding folder to FAQ list mapping, the FAQ list structure is expected to start at "Central" or the company level, and then follow the Hudu folder structure beneath that. This is an area we are open to revisiting, and feedback like this helps guide those decisions.

The original intent was for KB articles to be indexed by Halo’s AI. If Halo adds support for indexing FAQ lists as well, that would certainly be beneficial.

Thank you again for taking the time to share your input.

Oh....I've always wanted to go.

Was just looking at attending this for the first time

Appreciate this guys, think the 1st option is what am going to roll with for now. Now going to need another asset for Security Groups! Does this ever end!

Hey! Funny enough, I've had someone ask me this exact thing (and I just shared it with someone yesterday).

This is what I have:

Option 1 (what I would do):

1. Create an asset layout called “File Shares:”

   1. Create the following fields:

2. Your data would look something like this:

^^ I’d say that that’s probably the “best” way to handle it, but there’s no right/wrong answer. 

Option 2:

1. Create an Asset Layout called “File Sharing” 

   1. Create the following fields: 

      

It would look like this: 

For this second option, the actual data for SharePoint and the File Shares are just a table created in the HTML editor. The asset itself does link back to the parent server until the “Server” asset layout. 

Interesting idea. However, curious how this would hold up over time rather than just at creation.

• What happens when a process changes? If someone edits a step, does the workflow re-diff, warn, or require regeneration?

• How do you handle versioning and auditability once these workflows start doing real work? Being able to answer “what ran, when, and why” would be important.

• Is there a way to simulate or dry-run a workflow against sample data before letting it touch production systems?

• For non-automated processes, you mention "it should tell me up front so I stop wasting my time". A lot of processes may sit between manual and automatable, how do you envision handling these situations?

One other thought: a lot of people already stitch things together with n8n, Power Automate, or scripts alongside Hudu. I’d love to understand more clearly what your approach unlocks that those tools don’t, especially around trust, safety, and maintainability.

Overall, cool concept. If you can clearly define guardrails, human checkpoints, and failure behavior, I think you’ll get much stronger buy-in

P.S. you image does not seem to load on here, but I did see this over on Reddit as well (linked for others)

Hey Peter,

Using the Secure Note feature to send other items--I don't think it's really something intentional, rather something we've just never considered. Specific links can only be generated for KBs and Passwords right now.

Passwords, you can generate an auto-expire link to send to people (similar to how the secure note works).

KBs, you can generate a "public share link" to send them to anyone.

To share any other data, that would hinge on utilizing the "External Portal" (which would audit everything as well). Anything attached to KBs and assets are available via the portal as well.

Hi Alex Beauchamp and Daniel Gifford! Just to update -- spoke with Max and he was able to get this resolved with the community node 🙂

I added an issue to his Github page. See link: https://github.com/msoukhomlinov/n8n-nodes-hudu/issues/21

Jordan Hart -

The folder structure within Hudu doesn't sync quite right with Halo, and isn't flexible at all.

For example, I have Hudu organized at the top level as follows:

1 - Client Facing Documentation
2 - Internal Documentation
3 - Backoffice

The Internal Documentation is sync'd over 3 times? some articles are in one of them, some are in the other, etc. so I then I have to delete one, find the stray articles then manually put them into the other FAQ listing (folder) within Halo.

It would be AMAZING if we could 'map' folders to FAQ lists - even if at the top level. But if it would just sync properly without creating duplicates, it would be better than it is now.

Colter

I had this same issue. To resolve this, I deleted the API integration in HaloPSA and recreated it. It sync'ed everything in right away.

All of the categories sync'd in properly and from my checks so far the images are correct, etc.

I had a lot of trouble with this too as i needed the dropdown fields to correspond to the list items and also assign the asset to a user too. I ended up making an n8n flow for this as i gave up with the built in import solution.
It basically reads the .csv file, pulls in Hudu user assets and the lists needed for the asset layout and then a custom js node matched and prepared the API post calls.

Hi Cameron,

I appreciate you getting back to me! I will send you a DM right now :)

Hey there, Chris! Thank you for raising this question, as I'm sure you aren't the only person who uses company nicknames in such a way.

While nickname is not currently available as a direct query parameter (as shown below), you can still filter results client-side.

For example, when using the HuduAPI PowerShell module, you can filter companies by nickname (or any other property) using standard PowerShell patterns:

Get-HuduCompanies | where-object {$_.nickname -ieq "But folks call me Chris"}

Or for partial / case-insensitive matches:

Get-HuduCompanies | where-object {$_.nickname -ilike "*Chris*"}

This same logic can be applied regardless of the scripting language that you are using. If your project is using something specific, I'd be more than glad to help however possible!

I hope this helps!

Kind regards,
Mason

It can relatively easy be done with an automation script or tool. I currently use n8n and looking into creating a flow that adds all licenses and apps, with app registration and expiration dates, into Hudu assets. I know that the https://graph.microsoft.com/v1.0/servicePrincipals API holds the expiration dates.

Hi Chris! The most common way that I've seen is exactly how you described -- via an asset layout/expirations, but this is, unfortunately, a manual process. You could automate it, but would require utilizing the API.

If you're familiar with CIPP, I believe their integration with Hudu does have a feature for this.

I think there's a really neat opportunity for an automation library script or potentially including it in our M365 integration update (which is already planned). I've pinged my team to see if this is something they could look into 🙂

We do also already have a feature request for this: https://feedback.hudu.com/feature-requests/p/track-azure-app-registration-expirations