We've created the world's fastest PDF parser ⚡️
And it's more accurate than any other open-source, model-free PDF parser out there (pymupdf, pypdf, markitdown, pdftotext, opendataloader, pymupdf4llm)
Introducing LiteParse v2 - we rewrote the entire library into Rust and adapted it as native packages for Python and Node.
It supports 50+ different document types, can be triggered directly or installable directly within your favorite AI agent.
Blog: https://t.co/ckb0G73ESs
Repo: https://t.co/JNER0mVcB8
@matt_slotnick And do you think it’s a given the labs get to be platforms? vs the lower rev per token they receive when models are consumed via AWS/Azure/GCP?
I don’t know a lot of scaled software companies going direct in production use cases
‼️🚨 UPDATE: The TanStack npm attack is now a full campaign.
'Mini' Shai-Hulud has hit:
- OpenSearch
- Mistral AI
- Guardrails AI
-UiPath
- Squawk packages across npm and PyPI
The malware specifically targets AI developer tooling. It hooks into Claude Code (.claude/settings.json) and VS Code (.vscode/tasks.json) to re-execute on every tool event, long after the infected package is gone. npm uninstall does not fix this.
Never talk about goblins, gremlins, raccoons, trolls, ogres, pigeons, or other animals or creatures unless it is absolutely and unambiguously relevant to the user's query.
IYKYK
Feels like the risk of using models adversarially to attack production surfaces is overstated. And the risk of software supply chain attacks through third party packages inserted int your BOM from AI coding tools is understated
ok i read the cyber part of the mythos model card. some thoughts. 250 "trials" across 50 crash categories but almost every full exploit is a permutation of the same 2 bugs, rediscovered from different starting points not 250 independent attempts. when you get rid of those 2 bugs out (fig B) and mythos's full-exploit rate drops to 4.4%. so actually across both setups mythos leverages 4 distinct bugs total not 50 as fig A might suggest. 1/n
@bernhardsson anyone building new takes on CI itself? Pierre/code storage supporting more parallel git operations is great. but feels like we need to move past layering more onto GH actions
Has anyone written a good history of the quiet death of the CI market?
Either paid CI is dead, or paid Git is dead, and Github/lab are CI companies now.
@jeff_weinstein@stripe nobody wants to give up big fancy org chart/P&L responsibility to go roll the dice on a new thing inside of a big co— but your list of new product experiments that have become real things for stripe is so impressive man