Shelf

Every media tracker runs into the same wall: logging. Manual trackers ask for a diary entry for every film, a status update for every chapter, a rating for every album, and the person who started with good intentions has usually stopped updating by spring. Koodos built Shelf on the premise that logging should not be the user’s job at all. Connect accounts such as Spotify, Apple Music, Netflix, YouTube, Letterboxd and Goodreads, and Shelf assembles what you watch, read, listen to and play into a single profile that updates itself.

The result is part dashboard, part identity page. It surfaces top artists and genres, keeps a chronological history, tracks progress through books and shows, and delivers weekly and monthly recaps. A custom URL and background color push it closer to a personal homepage than a stats screen, and manual links fill in services the automatic connections don’t cover.

Shelf’s positioning is shrewd. It does not compete with the services people already use; it reads from them. Nobody has to abandon a reading app or switch streaming platforms to use it. That aggregation is the product’s main asset and its main dependency, because every feature is only as good as the connections feeding it.

Koodos also pitches a second identity: Shelf as a personal context app that learns a user’s interests over time, lets them review and correct that understanding, and shares only what they choose with other apps and AI agents. Community-built tools such as a weekly life recap and a release radar already hang off that idea. It is a much larger bet than a pretty profile. It turns consumption history into a portable taste model that other software can draw on.

Shelf holds a 4.6 out of 5 store rating from 4,095 ratings. That suggests the core profile lands; the open question is whether the personal-context layer can earn similar confidence. The integration risk is concrete: a failed service connection can leave a profile incomplete while still looking complete, and onboarding friction can block people before they reach any value at all.

• Automatic Aggregation: Connected accounts across music, video, film and books feed one profile, with no manual entry needed for supported services.
• Stats and Recaps: Top artists and genres, a chronological activity history, and weekly and monthly recaps turn raw activity into a readable picture of habits.
• Progress Tracking: Books and shows carry progress markers, so the profile reflects what is in progress as well as what is finished.
• Profile Customization: A custom URL, a background color, rearrangeable items and manually added links from other services give users control over presentation.
• Low-Pressure Social: Browse friends’ Shelves and react with emojis. There is no posting and no push to collect likes or followers.
• Private by Default: Privacy is on out of the box. Users choose which services connect and what others can see.
• Personal Context Layer: Shelf builds a model of a user’s interests that they can review and correct, then share selectively with apps, agents and community tools like a release radar.

The Passive Logging Loop

Shelf’s central design decision is to invert the usual tracker loop. Most trackers work on intent: you finish a film, open the app, search, log, rate. Shelf works on residue: you live your normal media life, and the record accumulates as a side effect. That removes the step where most trackers lose people. A diary that depends on discipline decays. A diary that depends on API connections only decays when the connections break.

The trade-off is depth for breadth. Automatic ingestion is good at answering what and when: which artists dominated the month, which show is half-finished, which genres keep coming back. It cannot capture why, and none of the documented features add reviews or written reactions. Shelf is a mirror, not a journal. For people who already keep detailed notes in Letterboxd or Goodreads, Shelf pulls from those services rather than replacing them, which is the right call. It lets dedicated users keep their existing habits while casual users get a complete record without effort.

The weekly and monthly recaps are the retention hook. They take the year-end streaming recap ritual and make it routine, and because they combine services, they show the cross-media pattern no single platform can: the podcast-heavy month that coincided with a reading slump, or the album that soundtracked a TV binge. That synthesis is the clearest argument for an aggregator, and Shelf builds its rhythm around it.

A Social Layer Without the Scoreboard

The social design is deliberately restrained. You can browse friends’ Shelves and react with emojis, and that is it. There is no feed of posts to keep up with, no follower count, and no like tally attached to anyone’s taste.

We think this is the smartest decision in the product. A media profile is a form of self-presentation, and attaching metrics to it invites performance, such as padding the history with prestige picks or hiding guilty pleasures. Without a scoreboard, the incentive runs the other way: the profile is accurate because nothing rewards curating it. Emoji reactions provide just enough acknowledgement to make browsing feel reciprocal without turning it into a transaction.

The low-pressure approach is coherent, and it gives browsing a calmer rhythm than a feed. The cost is reach. Without posting or discovery mechanics, Shelf’s social value depends on how many of your friends already use it. That makes it a slow-growth network by design, and one that offers a solo user little beyond the personal stats.

Privacy and the Personal Context Pitch

Privacy on by default is the correct starting point for an app that sees this much. A combined history of music, video, film and reading says more about a person than any one service does, and Shelf lets users decide service by service what connects and what others see.

The personal context ambition raises the stakes: Shelf is meant to learn interests over time and share them with apps and agents on the user’s terms. Two design choices make this defensible. The first is correctability: users can review and fix what Shelf thinks it knows, which matters because inferred taste models drift and misread. The second is selective sharing, which keeps the user as the gatekeeper rather than the product.

We remain cautious anyway. A correctable taste model only helps if people actually audit it, and most won’t. Granting access to agents is easy; understanding what an agent does with that access is harder. Community-built tools like the weekly life recap and the release radar show the upside clearly, since a release radar fed by real listening and reading history is useful. The downside is that every new consumer of that context becomes another party to trust. The framework is sound. The track record is still being written.

Where the Pipes Leak

An aggregator’s quality ceiling is set by its integrations, and the structural risk sits exactly there. A failed connection is the kind of failure that undermines the whole premise: a profile that silently misses one major service is an incomplete record that looks complete. Onboarding friction is worse in a different way, because it blocks people before they reach any value at all. A contacts prompt also sits in tension with the privacy-first pitch. It should be skippable without friction.

Neither issue is fatal, but both mark where Shelf is most fragile.

Shelf treats the profile as a canvas. A custom background color, rearrangeable items and a custom URL give each page a distinct identity, which matters for a product whose social value rests on browsing other people’s pages. Customization is central to the profile-as-canvas idea, though the palette can feel narrow; more color schemes would be an easy win for a product that is, in large part, about self-expression.

On performance, the meaningful variable is not the phone. Shelf’s heavy lifting happens through account connections and data syncing, so responsiveness and completeness depend mostly on third-party services behaving reliably. A stats-and-profile app of this kind should be light on device hardware. The real risk is integration reliability, and that is a problem of server-side plumbing rather than frame rates.

Shelf ships on iOS and Android, and both are touch-first experiences built around browsing profiles and recaps rather than heavy input, which suits a phone. The connected services are tied to accounts, not operating systems, so the core value of aggregation, recaps and friends’ Shelves should carry across both platforms.

The store rating is 4.6 out of 5 from 4,095 ratings, which points to a core profile that lands with users. The custom URL gives every profile a shareable address, which matters in a friend network that spans both platforms.

Shelf solves the right problem. Media trackers fail because logging is tedious, and Koodos removed the logging. What remains is a self-updating profile, regular recaps that connect habits across services, and a social layer smart enough to leave out the scoreboard. Privacy on by default and per-service controls show the team understands how sensitive a combined media history is.

The weak points are structural. An aggregator lives and dies by its integrations, and a failed connection or a stalled onboarding flow can leave the profile incomplete or block value entirely. The personal context and agent-sharing vision is ambitious and thoughtfully framed, with correction tools and selective sharing, but it asks for a level of trust the core profile does not need.

We recommend Shelf to people who want a low-effort overview of their media life and have a few friends on it; treat the agent features as promising rather than proven.