frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Show HN: tale.fyi, we deserve a home for fiction

https://tale.fyi/@sam/announcing-tale-fyi-read-or-listen-to-an-entire-book-from-a-single-link
31•samuelcole•1h ago•15 comments

Show HN: Formally verified 3D CSG: Trust 93 lines spec, not 1000 lines AI code

https://github.com/schildep/verified-3d-mesh-intersection
41•permute•1h ago•14 comments

Show HN: Ctrlb-decompose: Strip the noise from logs before sending to LLMs

https://github.com/ctrlb-hq/ctrlb-decompose
30•ruhani_grover•1h ago•3 comments

Show HN: Scala Tutorials – interactive Scala 3 lessons in the browser

https://scalatutorials.com
34•eranation•3d ago•4 comments

Show HN: Base-GPUI: A GPUI port of Base UI headless components

https://github.com/LukeTandjung/base-gpui
2•autodidacc•45m ago•0 comments

Show HN: Segue – Save context in one AI, load it in another by a short handle

https://segue.ai/
9•csaguiar•1h ago•5 comments

Show HN: Cloudrift – CLI open-source that finds wated AWS spend

https://github.com/elleVas/cloudrift
3•elleVas•58m ago•1 comments

Show HN: Citation Safe – MCP server verifies legal citations vs. CourtListener

https://citationsafe.com/mcp
3•gabera090•1h ago•0 comments

Show HN: PBTune – Evolutionary auto-tuning for PostgreSQL (no ML required)

https://github.com/alshawai/PBTune
4•alshawai•1h ago•1 comments

Show HN: Dinjure – Bulls and Cows in one HTML file, with a fallible AI

https://dinjure.com/
2•Dinjure•1h ago•0 comments

Show HN: Cosmos47, a public chronological feed without algorithms

https://cosmos47.com/
2•boredyeti•1h ago•2 comments

Show HN: SpecShield – Contract testing and can-I-deploy gates for OpenAPI

https://specshield.io/
3•DeepakSatyam•1h ago•0 comments

Show HN: Yap – OSS on-device voice dictation for macOS with no model to download

https://github.com/FrigadeHQ/yap
88•pancomplex•19h ago•30 comments

Show HN: BeatFlow: Codex skill for composing music as editable multi-track MIDI

https://github.com/the0cp/beatflow-skill
4•vaergawdd•3h ago•0 comments

Show HN: Kiln – a fast, native batch image processor

https://kilnimage.app
5•the-hangry-dev•4h ago•1 comments

Show HN: Hotcell – local sandboxes for AI agents

https://github.com/sinameraji/hotcell
2•sinameraji•2h ago•2 comments

Show HN: Open-source Cloudflare deployed agent native task management and wiki

https://tajd.github.io/projektor/
8•tajd•3h ago•0 comments

Show HN: How many people live within an hour walk of you?

https://blockatlas.com/tools/radius/walk
6•NameError•2h ago•2 comments

Show HN: BrowserAct: Browser Layer for Your AI Agent

https://github.com/browser-act/skills
5•aanthonymax•5h ago•0 comments

Show HN: The mentor setup I've run for months, now as a Claude Code plugin

https://github.com/hcsum/dont-let-me
2•sumtsui•2h ago•0 comments

Show HN: Zenly – Investment research platform built around a quality/value score

https://zenlyfinancial.com/
2•juliiii•2h ago•1 comments

Show HN: FeyNoBg – Automatic background removal model and training library

https://usefeyn.com/blog/feynobg/
114•snyy•21h ago•28 comments

Show HN: Fulltext search on PDFs and scanned images

https://github.com/Shadi/docs-search
2•_shadi•3h ago•0 comments

Show HN: Build an Avatar-Cursor Integration

https://lee-ai.com/
3•eschwalme•3h ago•1 comments

Show HN: We analyzed 18k flashcard reviews from our app's usage data

https://studycardsai.com/blog/we-analyzed-18000-flashcard-reviews
2•charliegroves•3h ago•0 comments

Show HN: Grassroots Football, Live

https://unleague.site/
2•kunalsin9h•3h ago•1 comments

Show HN: Trylle – The Next-Gen Git Platform for Modern Teams

https://trylle.com/home
17•Xlab•14h ago•10 comments

Show HN: Pre-registered trading backtests – 27 kills, 2 passes, 0 funded

https://strategygraveyard.com
2•Strawburry•4h ago•0 comments

Show HN: Dino Race – A peer-to-peer two-player Chrome Dino game

https://gameplayer2.com/dino
3•dares2573•6h ago•1 comments

Show HN: DiscountHub – AI-assisted search for deals and promo codes

https://discounthub.uz/
4•MasterBek•4h ago•0 comments
Open in hackernews

Show HN: Formally verified 3D CSG: Trust 93 lines spec, not 1000 lines AI code

https://github.com/schildep/verified-3d-mesh-intersection
41•permute•1h ago
To my knowledge, this is the first formally verified implementation of a 3D constructive solid geometry (CSG) operation: mesh intersection, implemented in Lean 4 and verified against a concise specification that pins down the surface of the resulting mesh exactly and guarantees practical well-formedness conditions on the triangulation.

This project is also an experiment in avoiding having to trust AI-generated code. A human reviewer only needs to read 93 lines of formal specification and run the Lean checker to certify the correctness of the kernel, skipping the intricate 1000+ lines of AI-written implementation. To prove correctness, AI autonomously wrote over 60,000 lines of Lean proofs, which also never have to be inspected by a human. The Lean checker guarantees conformance to the specification at compile time, with zero trust placed in any LLM. This allows us to treat the implementation and proofs as a black box. I guided the agent through the milestones described in the readme to arrive at the result presented here.

Also take a look at the web demo https://schildep.github.io/verified-3d-mesh-intersection/, which runs the verified mesh intersection kernel compiled to WebAssembly in your browser.

Comments

permute•1h ago
Note that the claims are about the kernel not the web demo: While the kernel is formally verified, the UI and glue code (that call the kernel) are not. I once hit a bug that looked like there were holes in the resulting mesh: I found that it was an overflow in the glue code (now fixed) that converts the exact rationals coordinates of the output mesh, for which we prove the specification, to floats before sending them to the GPU. All geometry is happening inside the kernel, for which we know the specification holds.
iFire•51m ago
How does it compare against https://github.com/elalish/manifold in performance and zero corruption?

I spent a lot of time making elalish/manifold work in Godot Engine and it is now a method in Blender too.

https://manifoldcad.org

If it helps, feel free to use the apache2 licensed unit tests that were generated in manifold development https://github.com/elalish/manifold/tree/master/test

iFire•45m ago
> Why we cannot have manifold output meshes in general

As far as I know manifold is not verified software, but it does promise that manifold inputs become manifold outputs.

I wonder how manifold will break in that condition or would it snap to the nearest manifold.

https://github.com/schildep/verified-3d-mesh-intersection#wh...

iFire•39m ago
On a readme style note, please spend more time explaining the algorithm and less time bragging on how amazing vibe coding is or how Lean has a zero trust proof.

I presume the new advance is claim that verified-3d-mesh-intersection is verified and not the claim that vibe coding can replace hand proofs or that the claim that lean4 can be correct without trusting llms.

permute•41m ago
The runtime performance is most probably worse than yours. The goal of this project is a minimal specification that a reviewer can trust. See section in readme on performance and how it could be improved while still keeping formal guarantees, but would mean a reviewer having to trust axioms on floating point numbers: https://github.com/schildep/verified-3d-mesh-intersection#pe...
bob1029•41m ago
This is really neat. I think CSG is one of the more intuitive ways to build 3d environments (especially 'indoor' ones). No geometry is ever destroyed in a proper CSG workflow. A hierarchy of brushes makes it possible to very rapidly iterate parameters like how long a hallway is or the time it takes to get from one objective to another. Often, you only have to apply a transform to a single brush to affect something that would take a non-CSG workflow an entire afternoon to recover from.

The current story with CSG around commercial game engines is pretty awful. Unity has some 3rd party options that can work well but they're also glitchy and don't integrate well with the other parts of the ecosystem. For a while I thinking about building my own CSG implementation but the exact things this project solved were what was stopping me. The EditorWindow integration stuff is not as scary to me (although it is painful). The advanced geometry and set operations is what kept me from trying. It appears that dragon has been (at least partially) slain.

iFire•38m ago
That's why I integrated manifold into godot engine for constructive solid geometry!

https://github.com/godotengine/godot-proposals/issues/9711

agentultra•40m ago
> A human reviewer only needs to read 93 lines of formal specification and run the Lean checker to certify the correctness of the kernel, skipping the intricate 1000+ lines of AI-written implementation. To prove correctness, AI autonomously wrote over 60,000 lines of Lean proofs, which also never have to be inspected by a human.

I am curious about the approach as I find claims like this hard to believe. There is a theory behind proof kernels that they must be small because they have to be trusted and verifiable by a human. Otherwise the whole system breaks.

How does one trust an LLM generated kernel is proving the right things?

permute•32m ago
Maybe what you mean is that kernels of proof assistants must be small. Here I am referring to a geometry processing kernel (that is formally verified by a proof assistant). The implementation of the algorithm can be very long, the proof that it conforms to the spec can be very long. But lean checks the proof. And so you only have to trust the spec and that the lean proof assistant is correct.

In the 93 lines I assumed a reviewer already trusts that the kernel of the Lean proof assistants is correct. We have to trust somethings.

CyLith•34m ago
The issue I have with all these formally verified numerical algorithms is that, at the end of the day, I need this implemented using actual floating point. It's great that it's verified, but not for the case that is practically meaningful. Unless it's implemented using hardware accelerated floating point, most of these algorithms are not sufficiently performant to be practically useful.
permute•27m ago
It is possible to prove things involving floats, you just have to addionally trust a small set of axioms on floats. See https://github.com/schildep/verified-3d-mesh-intersection#pe... Don't know if there is something like this in Lean. But there is https://flocq.gitlabpages.inria.fr
TacticalCoder•16m ago
> The issue I have with all these formally verified numerical algorithms is that, at the end of the day, I need this implemented using actual floating point.

You can go a very long way using discrete math though. I'm pretty sure it's doable to implement, say, a 3D slicer (to prepare layers/paths for 3D printing I mean) without using any floating point operation.

P.S: FWIW I was coding intros and demos using 3D effects on the Commodore Amiga and then on 386 PCs without using the 387 (the FP co-processor of the 386, when present). I'm not saying it's useful to play games in 4K: what I'm saying is that there are certainly applications where formally verified algos using integer math do make sense.

bartvk•30m ago
Congrats with this achievement. Do you see it possible that this project can be integrated into, for example, FreeCAD?
permute•23m ago
I think this would need to be reimplemented for performance. But maybe the spec can be reused. As agents get better we might reimplement the same spec for performance with minimal human effort. (See notes regarding performance and floats in readme.)