andrethegiant

I am so sick of these anthropomorphized names that have nothing to do with anything that we’re all supposed to remember now. Why are we giving products first names? The worst offender is probably Amazon Rufus. It’s all so dumb and I hate it. At least attempt to be clever and name it something that relates to the product itself. Even Google Wave, despite its shortcomings, made sense as a product name.

> I think there should also be a reasonable cap on the number of connections that can be made. Something like 300 friends sounds right. Any more than that and you're a collector, and not using the platform to foster connection.

Path[1] did that, but with a cap of 50, and then 150 (based on the Dunbar number of meaningful human connections one can retain). They had a crazy growth period but eventually went kaput.

[1] https://en.m.wikipedia.org/wiki/Path_(social_network)

The mitochondria is the powerhouse of the cell

Doesn’t this conflict with the original proposal of appending .md to any resource, e.g. /foo/bar.html.md? Or why not tell servers to respond to the Accept header when it’s set to text/markdown?

<link rel="alternate" /> is a standards-friendly way to semantically represent the same content in a different format

> If anyone has a browser plugin which automatically redirects to llms.txt sign me up.

Not a browser plugin, but you can prefix URLs with `pure.md/` to get the pure markdown of that page. It's not quite a 1:1 to llms.txt as it doesn't explain the entire domain, but works well for one-off pages. [disclaimer: I'm the maintainer]

Too busy rolling in money to actually write a substantial announcement?

Free tier (i.e. using API keys but without a paid subscription) is rate-limited to 10 requests per minute. https://pure.md/docs/#section/Rate-limits

> it seems to crash on any Arxiv PDF

Fixed, thanks for reporting :-)

Could be. But their pricing for the conversion is free, which leads me to believe LLMs are not involved.

Apart from lacking newlines, how is the result bad? It extracts the text for easy piping into an LLM.

Does it work on YouTube videos that have transcripts disabled?

Cloudflare’s ai.toMarkdown() function available in Workers AI can handle PDFs pretty easily. Judging from speed alone, it seems they’re parsing the actual content rather than shoving into OCR/LLM.

Shameless plug: I use this under the hood when you prefix any PDF URL with https://pure.md/ to convert to raw text.

Shameless plug: prefix any URL with https://pure.md/ to get the pure markdown of that page. Useful for direct piping into an LLM. Has bot detection avoidance, proxy rotation, and headless JS rendering built in.

Anything new here besides what was announced in March?

Always has been

If you like this, you'll also like https://root.vc

Buried in the article, a new CLI for coding:

> Codex CLI is fully open-source at https://github.com/openai/codex today.

Add a Cloudflare snippet / some other edge function, and transform the response to convert to plaintext

ChatGPT and others have some sort of attribution, where they link to the original webpage. How or when they decide to attribute is unclear. But websites are starting to pay attention to GEO (generative engine optimization) so that their brand isn’t entirely ignored by ChatGPT and others.

It’s a shame that the knee-jerk reaction has been to outright block these bots. I think in the future, websites will learn to serve pure markdown to these bots instead of blocking. That way, websites prevent bandwidth overages like in the article, while still informing LLMs about the services their website provides.

[disclaimer: I run https://pure.md, which helps websites shield from this traffic]

Why would a company go through the trouble of building a factory when the next administration could just reverse the decision four years from now? Building factories will take years

Would be interested in seeing a public repo if you want to make that available! :-)

Author will need to rewrite now that Llama 4 is out

What makes you think a twice impeached president who incited an insurrection would act in good faith?

https://github.com/puremd/puremd-mcp handles this, probably some other MCP servers out there that handle this too

Back then websites weren’t so resource intensive. The negative backlash towards bots is kind of a side effect of how demanding expectations of web experiences has become.

This is the MCP version of the pure.md[1] REST API. You can integrate it with MCP clients like Cursor, Windsurf, and Claude Desktop to help them reliably get context from web content.

If you'd like to use without configuring MCP, simply prefix any URL with `pure.md/` and you get the unblocked markdown content of that webpage. It avoids bot detection and renders JavaScript-heavy websites, and can convert HTML, PDFs, images, and more into pure markdown.

pure.md acts as a global caching layer between LLMs and web content. I like to think of it like a CDN for LLMs, similar to how Cloudinary is a CDN for images.

[1] https://pure.md/

You found the easter egg!