I predict that in a year or so, the way people interact with and think about Skills will be vastly different, in that we will barely know or think about them. Skills are the latest buzz in the world of AI trend catchers, preceded by prompt libraries and multi-agents, and they solve a really important problem, which is progressive disclosure of know-how, letting agents dynamically load context based on the task at hand. But just like their predecessors, they fail some basic first principles in the long run.

Take prompt libraries, or GPTs as they used to be called, why did they fail? No one wants to write prompts, let alone maintain and improve them. We just want the agent we are talking to do the thing we want it to do, the way we want it to do it (preferably even better than what we imagine). Same thing with multi-agents, who wants to sit there thinking about whether a task should spawn a sub-agent or not? I just want the agent I am talking to, to do the task I give it as quickly and as thoroughly as possible.

Skills have the same fundamental problem. If we look at what they get used for, it’s a handful of categories: sometimes they are a cache for anything an agent could find if it did some research (best practices for a React framework, how an API works, some source code analysis of a library), many times they hold user preferences on how they would like the agent to act, and most valuably they can capture some repeatable proprietary workflow that holds institutional knowledge (your company’s deploy process, the exact sequence of Jira transitions your team decided on). The first two are mostly filling gaps that better models won’t have, though there is still taste involved, choosing which best practices you subscribe to among many options isn’t something a model can assume on your behalf. But that’s exactly the kind of thing a good memory system should learn from working with you. I have Skills for different types of codebases, frontend and backend, and within those how I like the code to be for different languages, and that information fundamentally belongs in memory, not in files I have to manually curate. A model that has worked with me long enough to notice I always rewrite every resp to response and always reject inheritance for code reuse should just learn that, taste and all.

Most of the above use cases will be eaten by agents with better, more proactive memory systems, newer models, and agents that will focus on learning through questions first instead of trying to one-shot every prompt. If we think about where Skills end up from first principles, it comes down to whether the knowledge depends on the external world or on the user. External-world Skills are usually objective facts (even if they can be dynamic), things like how Svelte 5’s runes work or what the correct API is for a library that came out after the model’s training cutoff. Right now someone has to write these, keep them up to date, and figure out when to load them. In the future these will likely live in central repositories that agents search and pull from on their own, and as models get newer training data and better tool use, many of them will just stop being necessary altogether. User-dependent Skills are opinionated by nature, things like what makes a good PR description, how you like your code structured, what your team’s deploy process looks like. Right now you write these yourself, and you maintain them yourself, updating them every time your opinion evolves or your team changes something. In the future these will likely serve as kickstarters that morph over time through interaction, a starting opinion that the agent and user shape together until it reflects what the user actually wants, at which point it’s not really a Skill anymore, it’s a memory.

You might be hearing about how Skills are the hammer and every problem is a nail, and maybe there is some truth to that. But who actually wants to make hammers at home for every nail they see? If the AI competition keeps forcing Anthropic to innovate at the pace they are, I think it’s only a matter of time till they start pulling all necessary Skills to central places and giving their models first-class capabilities to load them, with the user only seeing that they were loaded. The rest will slowly turn to memories, and Skills will be a niche thing that is only kept for backwards compatibility.