Skip to content
PodSubs
ShowsPeopleGuestsJobsNewsEvents
More
BlogPodCred
Menu
ShowsPeopleGuestsJobsNewsEventsBlogPodCredNetworksSign in
Join nowSign in
  • Shows
  • People
  • Jobs
  • News
  • More
    • Guests
    • Events
    • Blog
    • PodCred
    • Networks
    Join nowSign in

Go High Level

Go High Level artwork

Test Your GoHighLevel Marketplace App — Complete Guide

Apr 25, 202623 minAudio

🚀 Start your FREE 30-day GoHighLevel trial: https://globalhighlevel.com/trial Learn how to test your draft and private Marketplace apps before publishing on GoHighLevel. This episode walks you through the testing process so you can ensure your app works perfectly before going live to agencies worldwide. In this episode you'll learn: • How to test draft and private Marketplace apps in GoHighLevel • Best practices for app testing before publishing • How to troubleshoot common app issues • Steps to prepare your app for marketplace distribution Ready to try GoHighLevel yourself? The link above gets you a FREE 30-day trial - double the standard 14-day trial. See why thousands of agencies run their entire business on one platform.

Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.

Transcript

From the show.

Imagine pushing just like a single line of custom code and instantly 50 of your clients live marketing campaigns just totally corrupt. Oh wow. Right. Automation's miss fire, leads disappear, and uh the entire database you promised to manage goes into a total meltdown. I mean, it is the ultimate digital marketing agency nightmare. It really, really is. But today, we are going to make absolutely sure that never happens to you. Welcome to the deep dive, by the way. Before we unpack exactly how to, you know, safeguard your agency's tech stack, I actually have a special welcome gift for you. Oh, nice. Let's hear it. Yeah, so if you check the show notes right now, you'll find a link for a free 30-day Go High Level trial. And that is actually double the standard trial length. Double the time. That's a massive runway. Exactly. So, if you've been uh waiting for a sign to jump in or you just want, you know, a huge runway to experiment with the platform, that link is waiting for you down below. And honestly, it is the perfect opportunity to implement the strategies we're looking at today. Because, well, we're diving deep into the architecture of custom development. Right. Specifically for agency owners. Exactly. How agency owners can safely build, test, and integrate custom apps into Go High Level. Because, you know, standard out of the box features are incredible. Yeah. But the ability to build bespoke white-labeled integrations. That's what gives an agency a truly unfair competitive edge. Oh, for sure. You can solve hyper-specific problems for your clients that generic software simply can't touch. Right. But, uh, the problem is that most of us are so used to the App Store experience. Yeah. You just tap a button on your phone, the app downloads and it just works seamlessly. Exactly. But when you step into the realm of custom API connections, you aren't just downloading a polished app, you are effectively wiring the electrical panel of your business while the power is like still surging through the building. That is a terrifying analogy, but it's so true. I mean, one wrong configuration in a live environment and you are literally playing with fire. Which is exactly why the documentation we're examining today is so critical. We've got the official Go High Level sandbox account guide, their app builder setup process, and a highly technical guest post by a developer, Sergio Leon, who breaks down OAuth flows. And reading through all of these, the one overarching theme across the board is just isolation. Yes, 100%. So, let's start right there with the concept of the sandbox account. Now, the documentation defines this as a non-production High Level environment specifically created for marketplace developers. Right. It's essentially a crash test dummy facility, a place to validate your app's behavior using mock data. Okay. But let me play devil's advocate here on behalf of the agency owners listening. Let's say I run a marketing agency, right? I already have a dummy test sub account sitting inside my main agency dashboard. Sure. And I use that to test out new workflows before I give them to clients. Why do I need to go navigate to a completely separate developer portal just to create specific sandbox? I mean, isn't that just extra bureaucratic red tape? It's a fair question. It really is. But using a regular sub account for development is a massive architectural risk. Really? How so? Well, when you test in a standard sub account, you are still fundamentally tethered to your production agency environment. So let's look at the mechanics of a webhook, for example. Okay. Webhooks. A webhook is essentially a notification system, right? When an event happens in High Level, it automatically pushes that data to your custom app. Now, imagine you write a bad line of code and your app receives that data, but then instantly triggers another event in High Level. Oh no. Which then sends another webhook. Exactly. Which triggers another event, which sends another webhook. So, you've just created an infinite loop. An aggressive, endless loop. Within minutes, you're firing thousands of requests a second. If you do that in a standard test sub account, you're going to eat up the API limits for your entire agency ecosystem. Wow. So, you could effectively DDoS your own business. You absolutely could. The sandbox accounts outlined in the sources are architecturally distinct. They are strictly quarantined to protect real customer data and preserve your live API limits. Okay, so it's a true quarantine zone. And looking at the documentation, it seems like High Level actually really incentivizes you to use this quarantine. Oh, big time. Yeah. These sandbox accounts include trial access to enterprise level features entirely free, just strictly for testing purposes. Right. So, if you want to see how your custom app interacts with the absolute highest tier functionality, you can do it without having to upgrade your live agency account just to run a simulation. Which is a huge money saver. That's a major benefit for developers who are validating marketplace apps or uh testing private integration tokens. Definitely. But we have to mention there is a very important operational constraint here. The ticking clock. Yes. Testing environments cost High Level real server space and compute power. So they aren't going to let you just park a crash test dummy in their facility forever. Right, the doc states the sandbox stays active for up to 6 months from the date of creation. Exactly. 6 months. So, you know, it is a generous window, but it is temporary by design. And they enforce that timeline, don't they? Very strictly. Yeah. After the 6 month mark, the environment may be deactivated. Now, you can request reactivation if you're legitimately still developing, but they will pull the plug much earlier if you violate their fair use guidelines. Because it's a testing ground to validate that a connection works. It is not like a load testing server meant to be bombarded with heavy traffic just to see what breaks. Exactly. Don't abuse the quarantine zone. Good advice. So, okay, once you have that quarantine 6-month environment established, the next logical step is designing the actual tool that will live inside it. And this brings us to the app builder interface. Right. The process starts simply enough. You know, you log in, go to your apps, and begin the creation process. But the very first decision the documentation forces you to make basically dictates the entire DNA of your project. The big fork in the road. Yeah. You have to choose between building a private app or a public app. And this is where enthusiasm often backfires for people. Yeah. A lot of developers, you know, they want to rush to the finish line, so they immediately try to build a public app. Because they want to get it listed in the High Level marketplace for the whole world to see. Right. Exactly. But the documentation explicitly advises against this. Always, always start as a private app. And a private app, just to clarify, is not listed in the marketplace. It is strictly intended for internal use or development. Exactly. Because a public app requires a really rigorous review process by High Level's team. I mean, they have to ensure your code is secure, your branding is compliant, and your integration actually provides real value to the broader user base. That makes sense. So, a private app allows you to bypass all that public scrutiny while you just iterate and break things in your sandbox. Right. You only flip the switch to public when the tool is rock solid. And honestly, for the agency owners listening, keeping it private might actually be the ultimate goal anyway. Because the sources detail a specific configuration for private apps that acts as a total game changer. Of visibility settings? Yes. When setting up the app, you are asked to define the target user and the visibility. And the recommendation is to set the target user to sub account and the visibility to both agency and sub account. Let's really unpack the strategy behind those specific settings because it's brilliant. Agencies operate by managing individual client locations, right? Which High Level structures as sub accounts. Right. So, by targeting the sub account, you're building a tool that operates at the individual client level, but setting the visibility to both the agency and the sub account allows for just incredible deployment flexibility. So, you as the agency owner can force install the app across your entire client base from the top down. Yes. Or alternatively, you can empower a savvy client to browse their own dashboard and self-serve the installation. Which is super cool. But I think the real magic happens when you pair those visibility settings with the listing type option. A white-label feature. Yes. The documentation highlights the white-label listing type. When you select white-label, the system basically strips away all the generic third-party developer branding. It wraps the entire integration in your agency's aesthetic. And the psychological impact of that on your client base really cannot be overstated. I mean, think about it. When your client logs into their portal, they don't see some cobbled together third-party plugin you bought off the shelf. Right. They see a highly specialized proprietary software feature that seamlessly matches the logo on their dashboard. It fundamentally changes the relationship. You stop being just, you know, a digital handyman who rents other people's software to run ads. You elevate your entire agency into a bespoke software provider. It creates incredible stickiness. I mean, if a client wants to leave your agency, they aren't just losing your marketing services. They are losing access to the custom software you built specifically to solve their operational bottlenecks. Exactly. Good luck finding that anywhere else. So true. Okay, so we've established the strategy and the branding. The documentation then breaks the actual workspace down into three distinct areas: build, manage, and insights. Right. The workspace trio. Yeah. The build section ham is the UI, the logo, the descriptions, and the marketplace modules that inject your app's features directly into the High Level interface. Then the insight section gives you visibility into performance metrics and installation counts once the app is live. Right. But sandwiched right between the face of the app and the metrics is the manage tab. And this, this is the plumbing. The plumbing, yes. Because you can design the most beautiful, intuitive interface in the world in the build section, but if that interface cannot securely communicate with the High Level database, well, you just have a dead web page. Exactly. Just a pretty picture. Right. The manage tab is where you configure the security architecture, which I know intimidates a lot of non-technical founders, but it actually follows a very logical progression if we break down the mechanisms. Let's do that. Let's start with scopes. So the API docs define scopes as the specific permissions your app is requesting. Basically, what data it is allowed to read and what data it is allowed to change. Yes. I always think of it like a corporate security badge. You know, when you hire a new intern, you program their key card to open the front lobby and the break room. You absolutely do not program their badge to open the server room or the CEO's office. Right. Exactly. Unless their specific job actually requires them to be in the server room. Right. The problem in software development is that programmers often just ask for the master key card on day one. Why do they do that? Well, they aren't entirely sure which specific API endpoints they'll eventually need to call as the app evolves, so they just go in and check every single scope box in the developer dashboard, mostly just to avoid dealing with access denied errors while they're writing code. Ah, I see. But the documentation explicitly warns against that lazy approach, right? They insist on the principle of minimum privilege. So, why is that master key approach so dangerous in a live environment? It comes down to a concept called the blast radius. The blast radius? Yeah. If your custom app is somehow compromised by a bad actor, or, honestly, even if you just write a disastrously buggy line of code, the damage is strictly limited to the scopes you authorized. Okay, so if your app only requested the scope to read calendar appointments, a hacker cannot use your app's credentials to, say, mass delete your client's financial records or export their entire contact list. Exactly. Minimal scopes strictly contain the blast radius. That makes total sense. It also requires you to justify your access to the end user. Like, when an agency client goes to install your app, they are presented with a consent screen listing exactly what you're asking to see. Right. And if a simple scheduling widget is demanding permission to modify their billing details, the user's going to instantly lose trust and just abandon the installation. Oh, absolutely. I would hit cancel so fast. Furthermore, if you ever do decide to transition your private app into a public app for the marketplace, the High Level security team will scrutinize your scopes heavily. If you're requesting access to data you don't actually need for the app to function, they will straight up reject your submission. Got it. Keep it minimal. So, we've defined what the app is allowed to touch. The next piece of the plumbing involves redirect URLs and webhooks. We touched on webhooks earlier with that infinite loop nightmare scenario, but let's clearly define the mechanism. Sure. A webhook is essentially a don't call us, we'll call you architecture. Instead of your custom app constantly knocking on High Level's server door every 5 seconds asking, hey, do we have a new lead yet? A webhook allows High Level to simply drop a data package at your app's doorstep the exact millisecond a new lead actually arrives. Which is vastly more efficient for server resources. Exactly. And then you have the redirect URL, which is a really critical component of the installation process. Because when a user approves your app, High Level needs to know exactly where to send the secret authorization code. Right. And the documentation is absolutely uncompromising here. The redirect URL must be an HTTPS connection. Must be HTTPS. Yeah. Because that means the transit of data between High Level and your server is fully encrypted. You cannot intercept it in plain text. Okay. And High Level requires absolute precision here. The URL you register in the developer dashboard must match the URL your server provides during the installation process down to the single character. I mean, even a missing trailing slash at the end of the web address will cause the entire security handshake to fail. Wow. Okay. So, attention to detail is paramount. And speaking of that security handshake, we really need to talk about the client key specifically. The client ID and the client secret. Oh, yes. These are basically the core credentials that authenticate your app's identity. And the documentation includes the stark, almost cinematic warning regarding the client secret. I'm actually going to quote it directly. It says, copy and store your client secret immediately. Once you click okay, it will not be shown again. Literally a burn after reading protocol. Seriously. It feels like a spy movie. It really is a brilliant piece of security architecture, though. When you generate that client secret, High Level displays the raw password on your screen just once. Right. The moment you click away, High Level takes that secret, runs it through a one-way cryptographic algorithm, and only saves the resulting scrambled hash in their database. Meaning, even High Level's own engineers cannot look up your client secret later. Exactly. The raw secret ceases to exist anywhere in the world except exactly where you chose to save it. So, if someone manages to breach your High Level developer account months later, there is no client secret sitting on the dashboard for them to steal and use to impersonate your app. That's incredibly secure, but I guess the downside is, if you lose it, there is no forgot password button. Nope. You have to generate a brand new client secret, which instantly breaks your app's existing connection until you go in and update your own back-end code with the new key. Ouch. So, bottom line, treat it like a master combination lock. Copy it, put it in a secure digital vault, and never, ever lose it. Okay, so now we have our keys. How do we actually use those keys to get access to client data? This is where Sergio Leon's guest post provides such incredible value. Oh, it really does. Because he breaks down the OAuth 2.0 protocol, specifically the authorization code flow that High Level uses. And look, OAuth can feel incredibly dense to read about, but Sergio simplifies it beautifully. It is fundamentally just a secure method of granting access without ever sharing a user's actual login password. Right. Now, the flow looks very different depending on whether your app is public or private. For public apps listed in the marketplace, High Level does all the heavy lifting. They handle the user interface, they provide the install button, and they manage the initial redirects. But our playbook today focuses heavily on building a private app. And Sergio points out that for private apps, the developer actually has to initiate the flow manually. You have to construct a very specific authorization URL yourself. This URL contains your public client ID, those specific minimal scopes we discussed earlier, and your precise redirect URL. And then you present that URL to your client. And when the client clicks that link, they're taken to a secure High Level login screen. They log in, they review the requested scopes, and they click approve. At that exact moment, High Level fires a temporary authorization code back to the redirect URL you provided. Right. But this code isn't the final key, though. It's just like a voucher. Correct. It's a voucher. Your back-end server takes that temporary authorization code, bundles it together with your highly guarded client secret, and sends both back to High Level through a secure behind-the-scenes channel. Okay. And if everything matches perfectly, High Level finally hands over the real prize. Yeah. An access token and a refresh token. And this brings us to what I think is the most dangerous technical hurdle for an agency owner. Because if you want your bespoke app to function flawlessly in the background without constantly bothering your clients, you have to understand the lifespan of these tokens. Absolutely. Because they are highly volatile. Sergio Leon lays out the reality check here. Access tokens, which are the actual keys used to pull or push data, they expire in exactly 24 hours. 24 hours. Once that clock hits zero, any API request your app makes will just be flat out rejected. So, your app needs a way to get a new access token without forcing the client to, you know, manually log back in and approve the app every single day, because that would be super annoying. It would ruin the user experience entirely. Right. So, that is the entire purpose of the refresh token. Your server quietly trades the refresh token for a brand new access token. Yes. But, and this is a massive, but the text reveals a serious booby trap built into the refresh token's architecture. Oh yeah, the booby trap. Listen to this specific condition. Refresh tokens are valid for one year or until they are used exactly once, whichever comes first. The self-destructing token. Exactly. The exact millisecond your server uses a refresh token to request a new access token, that original refresh token becomes completely invalid. It is dead. High Level processes the request and sends back the new access token along with a brand new refresh token in the response payload. And this is exactly where amateur integrations fall apart. Oh, completely. Wait, let me play this scenario out, because this seems incredibly fragile to me. What happens if uh my agency server sends the refresh token, High Level validates it, kills it, and sends the new tokens back? Right. But right at that second, my server drops the connection. Or, say, my database has a microsecond blip and fails to save the new refresh token. The old one is dead and I didn't catch the new one. Am I just permanently locked out of my own client's account? Yes. Wow. Yes. The cryptographic chain is completely broken at that point. Unless you have engineered a highly robust fail safe in your code to catch and store those tokens instantly, you are locked out. That is wild. And your only recourse is to go back to your client, explain that the premium custom software you promised them basically forgot who they are, and ask them to manually click the authorization link to start the entire process over. Which instantly shatters the whole premium white glove illusion we work so hard to build. I mean, you look like you don't even know how to maintain your own software. It destroys trust completely. This is exactly why token rotation logic is arguably the most critical piece of back-end engineering you will ever do for these apps. You have to prioritize the immediate persistent storage of that new refresh token the moment it arrives. So, why does High Level make it so difficult? Well, the reason High Level enforces this single-use rule is actually for your protection. Okay. How so? If a hacker somehow intercepts a refresh token and uses it to gain access, they consume it. Right. It's gone. Ah, I see. So the next time your legitimate server tries to use that same token, it will fail. And that failure acts as a massive alarm bell for High Level, signaling that a token theft has occurred, which allows them to automatically revoke all access and secure the account. Wow. So, it is brilliant security, but man, it demands meticulous engineering. It really does. You can't cut corners there. Okay. So, what does this entire architecture mean for you, the agency owner listening right now? Let's just recap the playbook for safely building your agency's tech advantage. Let's do it. First, you never test in production. You spin up a 6-month sandbox account to get pure isolation, protect your live API limits from those infinite webhook loops, and access enterprise features. Second, you always start by building a private app targeted at sub accounts. And you set the listing type to white label so the final product seamlessly blends into your agency's brand identity. Right. Third, you ruthlessly enforce the principle of minimum privilege with your scopes to contain the blast radius. You guard your burn after reading client secret in a secure digital vault. And finally, you engineer an air tight OAuth flow that meticulously catches and saves those self-destructing refresh tokens, ensuring your custom integration stays alive, secure, and running invisibly in the background day after day. It is a solid playbook. And you know, we started this deep dive talking about the sheer terror of breaking live client campaigns. But understanding the mechanics of sandboxes, scopes, and OAuth flows actually raises a much more provocative thought, I think. Oh, yeah. What's that? Well, now that you see how accessible and isolated this custom development process actually is, what if the ultimate competitive advantage for your digital marketing agency isn't finding the perfect software off the shelf? I love where this is going. Right. What if the real advantage is the hyper-specific, entirely bespoke private apps you build on top of High Level? I mean, you have the tools to analyze the exact nuanced bottlenecks your clients face and build the exact software required to solve them. Yes. You transition from being just a renter of generic software to the landlord of your own custom solutions. You become completely indispensable to your clients. That is a massive paradigm shift for an agency. And honestly, the absolute best way to start exploring that shift, to get your hands dirty in these sandbox environments and start blueprinting your first private app without risking a single piece of live data, is to take advantage of that special offer. Exactly. It gives you the necessary runway to architect everything we discussed today. It really does. So I will remind you one last time. The link for the free 30-day Go High Level trial, which, again, is double the standard time, is sitting right there in the show notes below. Go grab it. Yeah. Click it, set up your sandbox, start building your own white labeled solutions, and we will see you on the next deep dive.

Open the transcript file

Hosts and Guests

Nobody is credited on this episode yet.

Credit the people on this episode

Be the first to rate this episode

More episodes

All episodes
  1. Stop Chasing Leads: Auto Follow-Up in Conversation AI

    Oct 6, 202626 minAudio

    Your AI bot can now automatically follow up with inactive contacts - no manual work required. Learn how HighLevel's Conversation AI detects drop-offs, schedules smart reminders, and closes deals on autopilot. Watch this...

    Credit the people on this episode

  2. Knowledge Base Triggers: Control Your Conversation AI Responses

    Oct 6, 202617 minAudio

    Stop letting AI guess when to show offers and pricing. Learn how GoHighLevel's Knowledge Base Triggers deliver the right content at the exact moment your prospects need it - driving more conversions and eliminating...

    Credit the people on this episode

  3. Conversation AI on Mobile: Reply to Customers Instantly (No Desktop Needed)

    Oct 6, 202617 minAudio

    Manage customer conversations from your phone with Conversation AI in the HighLevel mobile app. Learn how to use Autopilot Mode, Suggestive Mode, and control your AI assistant on-the-go across WhatsApp, Facebook,...

    Credit the people on this episode

  4. Voice AI Agent Transfer: Route Calls Without Dropping Them

    Oct 6, 202623 minAudio

    Learn how GoHighLevel's Voice AI Agent Transfer hands off live calls between specialized agents seamlessly - no dial transfers, no restarts. Discover how to route callers to sales, support, billing, or language-specific...

    Credit the people on this episode

  5. Sell Your AI Agents: GoHighLevel Marketplace Launch Guide

    Oct 6, 202620 minAudio

    Turn your Conversation AI and Voice AI templates into revenue streams. Learn how to package, list, and sell your agents on the GoHighLevel Marketplace - complete with workflows, automations, and knowledge bases...

    Credit the people on this episode