Real talk. Everybody's stuck on something. Vote below so we can see where this group actually is. If you pick "Other," tell us what it is in the comments. What's your #1 blocker this month? I'll post the results Saturday, and the top answer gets a how-to thread next week. Johnny

Robb do you have a cold email or cold prospecting system? just curious

A workflow list with forty items called "New Workflow" is a trap. A simple naming habit helps: start with the client or area, then the trigger, then the goal. For example, "Dental | Missed Call | Text-back". When something breaks at 9pm, you'll find the right one in seconds instead of opening five. What's your naming convention for workflows?

I previously posted about my GoHighLevel (GHL) MCP connection constantly disconnecting from Claude and ChatGPT, forcing me to reauthorize it every few days. After investigating, I discovered that HighLevel now has newer, client-specific MCP V2 endpoints that I wasn't using. I was still connecting through the original legacy endpoint: "https://services.leadconnectorhq.com/mcp/" Apparently, HighLevel has introduced dedicated MCP endpoints: For Claude: "https://services.leadconnectorhq.com/mcp/anthropic/v2" For ChatGPT: "https://services.leadconnectorhq.com/mcp/openai/v2/" According to HighLevel's documentation, these newer connections offer: - Dedicated OAuth authentication for each AI client. - Refresh-token support for longer-lasting connections without repeated manual authorization. - Expanded access to GHL CRM operations. - Better support for managing sub-accounts and permissions. What I believe was happening: My legacy MCP connection may not have been maintaining authentication properly, causing Claude and ChatGPT to repeatedly request reauthorization. I haven't confirmed the exact root cause, but the older connection was clearly unreliable in my setup. The potential fix: 1. Remove or disable the legacy MCP connection after setting up a replacement. 2. Add the appropriate V2 endpoint in Claude or ChatGPT. 3. Reauthorize through GHL OAuth. 4. Select the correct GHL sub-account and permissions. 5. Test whether the connection remains active over several days. I'm switching to these newer endpoints now, so I can't confirm yet whether this permanently fixes the disconnections. Official GHL documentation: https://marketplace.gohighlevel.com/docs/other/mcp/ Hopefully this helps others dealing with the same issue. Has anyone already migrated to MCP V2? Has it actually solved the frequent reauthorization problem for you?

When SMS stops landing, the workflow is often fine and the cause is somewhere else. Start with the registration status of your sending number, since unregistered or pending numbers get filtered. Then look at the message itself: shortened links, all caps and heavy punctuation get flagged faster. Check that the contact actually opted in and is not marked do not disturb. Look at quiet hours and the contact's time zone, because a message can be held or sent at a bad hour. Last, check the conversation log for the exact delivery status of one failed message before changing anything. Working through those in order usually finds it in a few minutes.