Our MCP server is entering BETA, so I’m starting this thread to document different use cases, prompts, and things I learn along the way. I posted the initial MCP update here: https://www.skool.com/leadcapture/mcp-just-hit-staging-soon-agents-will-be-able-to-interact-with-your-leadcapture-account?p=324d6a03 It’s pretty mind-blowing to watch an AI agent interact directly with LeadCapture, review account data, analyze funnels, and make changes. The analysis side in particular is opening up use cases I hadn’t even considered. An agent can slice and dice your account data in ways that would normally take a lot of manual digging. I’m going to use this thread as a running prompt dump for things that work well. Some MCP use cases I’m testing: - Clone a funnel and make specific edits - Set up a routing rule - Build a lead quiz from scratch with X questions - Change styling across all funnel steps - Review all funnels and tell me which had the lowest conversion rate over the last 5 days - Audit my entire LeadCapture account - Import Meta or Google Ads conversion data and compare it against LeadCapture performance - Identify funnel steps with unusually high drop-off - Compare conversion rates across funnels, traffic sources, or date ranges - Find opportunities to improve conversion rates based on account data ⚠️ Important BETA note Getting an MCP server right is surprisingly difficult. Right now, MCP is strongest for analysis and account intelligence. We’re still expanding the tools it can access for editing, building, and making changes inside LeadCapture. That’s currently the biggest gap, so some prompts that involve actually changing your account may not work yet. That’s also the purpose of this Beta. We want to see what people actually try to make agents do so we can keep expanding the MCP tools around real-world use cases.