So I'm showing here an automation that I had previously made on its own just by letting claude code coach me on how to solve my manual report pulling loop for a custom membership churn dashboard we made. It doesn't use Trigger.dev but it runs some playwright scripts to pull the reports each morning and upload them into my dashboard. then these indicators show if we have a failure or not. Just went through my the first round of troubleshooting after the website I'm using to generate the reports went through a UI update. I can see how using something like trigger dev or having any sort of indicator so that failures don't say silent or must have!

Here's one of my early skills, a simple email brief. Like a lot of people I spend time catching up on emails because they pile up so fast. So I made a brief skill that sends me a summary of basically High priority, Medium, and low/skippable emails. One adjustment I made was setting up message forwarding because I actually have two work emails and then adding that context to the skill so it knows everywhere to check. Then I was able to pull the personal parts of my context out of the skill to make a team-wide skill where they can fill in their individual role goals and things that are priorities for their skill, based off of the same scaffolding that's at the core of this brief so they get like an 80% head start on an email brief skill!

So I don't really have a great screenshot of this but I built a revenue tracking tool for one of the golf course properties we manage. This tool ingests multiple reports that are pulled from the point of sale so that it has the most accurate data we can get. I went ahead and asked Claude Code if we can use an MCP or an API to connect it directly to this revenue tool so we can begin asking plain English questions about our business and its trends, and it worked! Our Team now has direct access to our transaction reports, special code usage, and membership reports so we can get some pretty comprehensive breakdowns of business trends, if a promotion is working or not, and overall what's changing year over year, MoM, or in our trailing 7 days! Then Claude can act as a sparring partner while we gameplan what's working and what's not working. #AISChallenge

Not because it was broken. Because it made the product worse. 😭 I built this whole feature thinking: "People are going to love having more control." Then I watched someone use it. They got confused. Clicked around. Ignored half of it. And eventually asked: "Which option am I actually supposed to choose?" That hurt. Because the feature worked exactly as designed. It was just solving a problem nobody had. So I removed it. The product instantly felt cleaner. One thing I'm learning: The hardest part of building isn't adding things. It's knowing what to kill.

Our business operates multiple driving ranges that run the same golf sim technology. Our owner loves to run multi site contests but gets stuck spending hours each week logging scores manually and sending updates. Spun this up with all the rules he put together and he was pumped! For future projects like this I'll definitely take some extra time in planning mode so we get a more complete product in the beginning. #AISChallenge