Introducing Wikiwise: an open-source Mac app for managing your own Karpathy-style LLM wiki. Set up a new wiki in a few clicks: all you need is Wikiwise + your agent. It's infinitely customizable, just markdown/html under the hood, and one click to share your wiki publicly. Here's how it... works: * Install Wikiwise for mac (it's built in Swift so super minimal and performant). In Karpathy's framework, Wikiwise is your IDE. * Start a new Wiki: it generates a new folder on your machine that's scaffolded in the wiki structure Andrej Karpathy describes (index.md, raw folder, wiki folder, CLAUDE.md/AGENTS.md, although it tries to be as un-opinionated as possible). * Then just point your agent (Codex, Claude Code, Cursor, etc) at the folder and tell it what to import -- files on your machine, connect to your Readwise account, or urls from the web. * Your agent creates wthe wiki for you: Your agent will know how to ingest your raw sources (via the AGENTS.md) and will immediately start writing+linking wiki pages for you. * Go crazy on customization! The rendered wiki pages live as static html/css in your folder too so just tell your agent to change stuff, and if you need any more customization Wikiwise is fully open source :) * Ask questions about your research with your agent, ask it to bring in new sources, write new documents, etc. * (optionally) Hit the Publish button to share your wiki with friends/colleagues at a custom URL === I tried to walk the line on a couple constraints with Wikiwise: 1. I wanted it to be easy to spin up new wikis, especially without chaining together a bunch of different apps. It takes me a few minutes to spin up a new wiki on a topic -- I already have five! 2. Infinitely Customizable: one great aspect of building a wiki as Karpathy described is that you can modify any aspect of your wiki with your agent. Every new wiki styling+structure is self-contained in the local folder, which allows you to preserve this. Wikiwise is just an IDE that makes the setup easier and includes a nice un-opinionated starting state. 3. Minimal: Wikiwise is built mostly in Swift, and the DMG you install to download it is only 2.6MB (!) 4. Easy Publishing: my colleague Eleanor Konik has been building her own LLM wikis for months, but has always really struggled to actually share them with her book club. There are tools to do it, but figuring out hosting is always a huge headache. This seemed like an ideal usecase for a tool like Wikiwise to solve. The process of building wikiwise was also pretty interesting -- I "bootstrapped" the app in a way by first building my own wiki based on Karpathy's tweet and other notes I had, and slowly formed the shape of the project in collaboration with my LLM. This was all done in 3 days over the latest Readwise company hackathon we had. Truly an incredible time to be alive. Anyways, curious what you think! Links in next tweet.show more

You can install Wikiwise here: View the source on Github: View the Wiki I created in the video here: or the other wiki I mentioned here: LMK what you think :)

hahahah yesssss that was the inspo, thanks to @deadly_onion

Thanks for sharing 🙏🏼

this is awesome, can’t wait to give it a spin. putting those readwise highlights to good use

exactly! so much you can do with them

do you have a favorite use case?

Love this! Building the same thing with

haha cool, from performative notes to performative wikis 😂

Lol The goal is to have people learn more from their notes via wiki. Lets hope wikis wont be just sitting there

this is fascinating - have been using Cowork over Obsidian but basically this is what i'm trying to do! and more integrated

nice! i can imagine cowork is great for this kind of stuff :) you could probably point cowork at a wikiwise directory too instead of a terminal agent!

can we make a pf and send a donation to your github? would love to support this bro.

THAT IS SO COOL!!!!!!

This looks great! I'm building something focused purely on the publishing side at I find LLM notes most useful when carefully assimilated into my own densely-linked notes. Publishing meaningful extracts from those connected notes requires a lot of care.

@readwise this is awesome Tristan! Loving how much cool stuff you are shipping

@readwise Thanks! This was just an experiment at our team hackathon but i think the direction is super interesting!

Looks great, just need to buy a Mac!!

😭 "claude make a windows wikiwise app. make no mistakes"

🤣🙏

Thank you - this is the only acceptable Obsidian alternative I will support!

Super cool! Really curious what’s the intended use and the applicability of the graph view experience ?

You can use it for any kind of research, notetaking, learning, etc. I'd recommend checking out the quoted tweet from @karpathy , it's a way of implementing his "LLM Knowledge Base"

I wrote more about this below, but I still don’t see graph view itself as the real day-to-day value. Visualizing connections is cool to play with. The real unlock is when an LLM can use the underlying KG, traverse those relationships, and turn them into useful output🤔

@karpathy ohhh you just mean the graph view yourself? i don't think it's a big deal in terms of these wikis, it's just a fun way to get an overview of the connections and pages the LLM can already traverse the links so it can do that!

@karpathy Right exactly. What feels compelling to me isn’t graph view by itself, but graph view behind the output: you work with the LLM on something in canvas, and as it evolves, you can see the knowledge map it’s pulling from and connecting through.

COOOOOOK

built a simple npx too scffold your obsidian vault & qmd - npx claude-second-brain - thats it -

that's really cool! but obsidian on its own: * is not open source, nor customizable enough * is cluttered -- does a lot of things and has a lot of functionality/core plugins you don't need * publishing is pretty brutal UX + $$$

No arguments there. Obsidian distribution and brand recognition is unbeatable. Everyone in tech has tried it at some point in time. So felt like if the plumbing was faster/low effort with familiar tools, people might be more likely to give it a shot

We just open-sourced OpenKB: Open LLM Knowledge Base In addition to Andrej's original design, OpenKB can scale to long PDFs and multi-modality.

I will put all my personal files into a second-brain app some random dudes vibe-coded last Sunday night? NYET!!!

Look cool but I know people for them less complicated matters, and Andrei did his way really good for it

Read AGENTS.md. The "edit 2-3 related pages" rule in the ingest skill got me thinking. Curious if it scales past 50ish pages, or if the agent starts grabbing the obvious links and missing subtler ones. Going to spin one up this weekend and see. Have you seen recall fall off as wikis grow, or does lint usually catch it?

Good question! I haven't seen any recall issues, but I also haven't gone past 100 pages. To me that goes from a topic-specific wiki (which karpathy originally described) to a general wiki... also a cool concept but a bit of a different problem. That being said, I think it's pretty trivial to just extend the AGENTS.md for your wiki to scale better.. and of course can add some things like qmd if needed!

nice, makes sense. I was thinking who does topic clustering for readwise but just saw the import-readwise skill and that readwise cli handles it. clean.

"one click to share publicly" is the interesting design choice here — most wiki patterns assume private knowledge compilation. making it public-first changes the incentive: you curate differently when you know others will read it. that tension might actually improve knowledge quality.

how is this compared to llmwiki? I am curious

You can think of this as an IDE for implementing+browsing+publishing your own llm wiki!

A terminal-bound relic shaped by Andrej Karpathy’s vision hoarding fragments of knowledge like blood in sealed vaults, waiting to be drawn.

Something to keep in mind: Claude Code requires a Claude Pro/Max subscription or Anthropic Console API credits because every time you ask it to ingest a file, you are using Anthropic’s high-end cloud computing power.

This might be the cleanest take yet on Andrej Karpathy’s LLM wiki idea, turning notes into a living system. Saw something similar done by a MIT prof, seems inspired by karpathy:

We've actually made a similar format in terminal based of Karpathy's idea of LLM Knowledge Bases. Give it a spin and let me know what you think