I’m a big fan of ClickUp.
I’ve used it for a couple of years now and I find it genuinely useful for building custom tables, tracking projects, managing to do lists and keeping information in one place without it disappearing into the usual digital fog.
One thing I’d not managed to get right, though, was a simple CRM-style workflow for capturing contact information from LinkedIn straight into ClickUp.
That was the gap.
I could already see other tools doing a version of this. OnePageCRM, for example, has a Chrome extension that lets you capture LinkedIn contact details and push them into its own system. Useful idea. Sensible product. I just didn’t want to pay for another subscription to solve a fairly narrow problem when I was already paying for ClickUp and wanted the data to live there anyway.
So this morning I tried something different.
Rather than looking for another tool, I asked Claude Code to help me build the functionality I actually wanted.
The problem was simple, just annoying
My workflow need wasn’t especially exotic.
When I’m on someone’s LinkedIn profile, I want to capture the useful details and move them into ClickUp as a potential contact for future follow-up and lead nurturing. Profile link. Current role. Company. A few key pieces of information that help me remember who they are and why they matter.
Nothing dramatic.
The irritating part was the manual process. Open LinkedIn. Copy the name. Copy the profile link. Copy the company. Copy the role. Jump into ClickUp. Paste everything into the right fields. Realise you missed one. Go back. Repeat.
That kind of friction is rarely a strategic problem. It’s just enough of a nuisance to slow you down, and just repetitive enough to make you wonder why you’re still doing it by hand.
So I asked for a spec, then a build
I didn’t start by asking Claude Code to magically produce a perfect finished app from thin air.
I asked it to create a design spec for the functionality first. I wanted a Chrome extension that could work from a LinkedIn profile page, capture the relevant information, and send it directly into my ClickUp environment. I also wanted to be able to choose the right ClickUp list and map the fields properly so the data landed where I wanted it.
That part matters more than people think.
The longest part of the whole exercise was not the technical build. It was deciding what fields I actually wanted in my ClickUp table. Once I got clear on that, the rest moved quickly.
Very quickly, actually.
Claude Code generated the spec and then got to work on the extension itself.
The setup was surprisingly straightforward
The ClickUp side was easy enough. I needed an API key, which took a couple of clicks to get from within my ClickUp settings. I copied that into the extension and, importantly for me, the key stays on my machine rather than being pushed into some hosted environment I didn’t control.
From there, the extension enabled me to:
- connect to my ClickUp environment
- find the right list
- map the fields exactly how I wanted them
- capture the relevant details from LinkedIn
- send that information straight into ClickUp without the usual copy and paste routine
Within about 15 minutes, I had a working version of the core functionality.
That still feels slightly ridiculous.
It wasn’t perfect first time
Worth saying, because the fairy tale version of AI tends to miss this part.
It didn’t get everything right on the first pass.
One of the fields in my ClickUp table had been missed in the original mapping, so I had Claude Code tweak that. There were also a couple of small layout adjustments I wanted in the extension. Nothing dramatic, just the normal kind of iteration you’d expect when you’re turning a rough first version into something you’ll actually use.
But that was the point. These were tweaks, not a rebuild.
I wasn’t stuck in a cycle of writing a long brief, waiting days, reviewing a mock-up, then going back for another round. I was making small corrections in real time and getting to a usable version quickly.
That changes the feel of the whole thing.
The real value wasn’t just the time saving
Yes, it saved time.
Yes, it saved me money because I didn’t take out another subscription for functionality I could now replicate inside a system I already use.
Yes, it’s made contact management cleaner and more practical in my own environment.
All true.
But a year ago, I probably wouldn’t have attempted this. I’d have assumed I had three options: live with the inefficiency, buy another product, or ask a developer to build something custom.
This morning, there was a fourth option. Define the workflow clearly, use AI to create the spec, build a first version, make a couple of tweaks, and get on with using it.
That’s a very different set of economics for small software problems.
The bottleneck is moving
That’s what I find most interesting.
For narrow, useful, web-based workflows like this one, the limiting factor is increasingly not “can this be built?” but “am I clear enough about what I want built?”
That’s a different challenge altogether.
In my case, the hardest part was being specific about the fields in ClickUp and how I wanted the information transposed into the list. Once that was clear, the actual build moved at a pace that would have seemed slightly absurd not that long ago.
And I don’t say that as someone trying to sound like a developer. I’m not.
I say it as someone who understood the workflow problem well enough to describe it properly.
There’s something quite important in that. The people who understand the operational pain points, the repetitive tasks, the clunky handoffs and the awkward workarounds are now much closer to being able to solve them themselves.