400 saved, 12 read - what I built after Pocket died

When Mozilla announced Pocket was shutting down in May 2025, I had four hundred and seventeen saved articles in my account. I'd read twelve of them.

Twelve total. Across three years.

I had three months before the lights went off - Pocket's apps came down on May 22, 2025, the service died on July 8, and Mozilla deleted everyone's data on October 8 - and the migration tools were generous. You could export everything to Reader (Readwise's renamed app), import to Raindrop, push to Instapaper. I tried all three. Six months later, I had three new graveyards instead of one.

The pattern was identical every time. Install the tool. The saving gets easier. I save more. The pile grows faster than I clear it. Eventually I declare bankruptcy and migrate to the next thing, which becomes a new graveyard within six weeks.

It took me longer than I'd like to admit to notice the pattern. Every read-later tool on the market - the ones that survived Pocket and the ones that built on its corpse - is solving the wrong problem.

They all make saving easier. None of them help you finish.

Saving is the easy part. Saving feels like progress. You hit the button, the article disappears into a tidy interface, and you get a little dopamine hit of "I'll handle that later." But later never comes, because the next interesting thing is already in your face, and the pile has no resolution mechanism. There's no equivalent of inbox-zero. The only way most people clear their pile is to declare bankruptcy on the whole list and start a new one - which I did, twice, before Mozilla made the decision for me.

Pocket dying didn't fix this. Reader, for all its highlighting and AI features, is still fundamentally a save-it-and-feel-better tool. Raindrop is a beautiful bookmark manager that has the exact same problem. Instapaper has been around the longest and has clearly given up on solving it. Every alternative I've tried sells the saving, polishes the interface, adds a new ranking system or a smarter feed, and stops there. The pile keeps growing.

So I stopped trying to read the pile. I started trying to resolve it instead.

The distinction matters. Reading is high-cost: it requires sustained attention, the right mood, the right time of day. Most saved articles never clear that bar, which is why they sit forever. Resolving is different. To resolve a saved article, you only need to answer three questions: Is this still relevant? Is it worth the time? What's the takeaway if I don't read it in full?

That last question is the one nobody builds for. Every existing tool either gives you the article or a generic summary that's basically the article with the verbs removed. Neither is useful. What I actually wanted was an action brief - a 60-second read that told me what the piece argues, what it's worth, and what to do about it. Read fully. Skip. Save the takeaway. Close the tab.

So I started building one.

The first version was a Chrome extension that ran any saved page through a fast model and returned three bullets and a one-line verdict. Crude, but it worked. I cleared 80 of my 417-item pile in one weekend, mostly by skipping. About 40 of those skips were articles I'd been guilt-carrying for over a year. I didn't miss them. The world did not end.

The second version added what I now call triage mode - a swipe-style interface that hands you one action brief at a time and asks you to decide. Skip, read fully, or save the takeaway. The whole pile becomes a stack of cards that you can clear in a focused 20 minutes instead of dreading for months. It's the single feature that turned this from "another summarizer" into something that actually changed how I relate to saved content.

The third version added a daily skim - one digest in the morning of what I saved the previous day, ranked by what's most likely worth my time. Five things, three minutes. The first read-later product I've ever opened more than once.

I named it skiml. The name is on purpose: the things you save deserve a skim before they earn a full read. Most don't earn the full read. That's not a moral failing on your part - that's the math of how much gets published versus how much one human can read. The job of a serious tool in this category isn't to help you read more. It's to help you decide faster.

If you were a Pocket user who's still not happy with where you landed - if Reader feels too heavy, Raindrop feels too much like a bookmark manager, and you're still carrying around the same guilt-pile you had before Mozilla pulled the plug - you're who I'm building this for. The pile isn't your fault. The tools have been wrong. There's a way out.

skiml is a Chrome extension and a dashboard, free during the beta. The waitlist is open at skiml.io. I'm building a feedback loop directly into the dashboard so the first cohort can shape the product as it grows. The beta is small on purpose - I'd rather have 50 people who actually use the thing than 5,000 who signed up and ghosted.