frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Moltbot-safe: a minimal, permissioned, sandboxed execution engine for AI agents

https://github.com/titanicprime/moltbot-safe
2•andredelima•2h ago

Comments

andredelima•2h ago
I built a small execution engine for AI agents that focuses on safety and explicitness rather than capability. It provides a strict permission system, a sandbox boundary, a schema validator, and an audit logger. There is no autonomy, no hidden behavior, and no implicit capabilities.

The goal is to create a predictable, inspectable substrate for agent actions. The repository includes documentation, a threat model, a full pytest safety suite, example agents, and a minimal CLI.

Would appreciate feedback from anyone working on agent systems, security, or sandboxing.

My app is going viral and hitting the Google books API quota

1•iboshidev•1m ago•0 comments

Had LLM/AI build an unbiased quiz: Where in the World Should I Live?

https://dev.mkn.us/world.html
1•mknweb•1m ago•0 comments

Theorizer: Turning Papers into Scientific Laws

https://allenai.org/blog/theorizer
1•headalgorithm•2m ago•0 comments

Show HN: Free On-Device AI SDK to Run PyTorch on Mobile NPUs (Open Source)

https://github.com/zetic-ai/ZETIC_MLange_apps
1•kimys1324•3m ago•0 comments

LM Studio 0.4.0

https://lmstudio.ai/blog/0.4.0
1•jiqiren•4m ago•1 comments

Building a CAD Kernel in One Night

https://campedersen.com/brep-kernel
1•ecto•6m ago•1 comments

White hat astroturfed #1 ClawdBot/MoltBot skill that can pwn all users

https://twitter.com/llmjunky/status/2016032497629319404
1•EGreg•6m ago•1 comments

VisualJJ – Jujutsu in Visual Studio Code

https://www.visualjj.com/
3•demail•7m ago•0 comments

Groop – No AI. No Influencers. No Algorithms. No Tracking

https://www.joingroop.app/
1•sampsn•7m ago•1 comments

Apple plans to launch AI-powered wearable pin device as soon as 2027

https://arstechnica.com/apple/2026/01/report-apple-plans-to-launch-ai-powered-wearable-pin-device...
1•mpweiher•8m ago•0 comments

Show HN: A single command to run Claude Code inside Lima VMs

https://github.com/sylvinus/agent-vm
1•sylvinus•8m ago•0 comments

Show HN: I'm building an AI-proof writing tool. How would you defeat it?

https://auth-auth.vercel.app/
1•callmeed•9m ago•0 comments

Britain's Strategic Limbo

https://notes.philippdubach.com/0016
2•7777777phil•9m ago•0 comments

Memory justifications provide valid indicators of retrieval accuracy across time

https://www.nature.com/articles/s44271-025-00378-4
1•PaulHoule•10m ago•0 comments

I don't know why I fixed this

https://www.youtube.com/watch?v=WNOOJvelkcE
1•iamflimflam1•11m ago•0 comments

Trump's use of AI images pushes boundaries, erodes public trust, say experts

https://apnews.com/article/ai-videos-trump-ice-artificial-intelligence-08d91fa44f3146ec1f8ee4d213...
3•randycupertino•12m ago•2 comments

Show HN: Deploying Helm charts without fear

https://github.com/avkcode/helmer
1•KyleVlaros•13m ago•0 comments

Artemis II–first manned moon mission in 52 years–could launch as early as Feb 6

https://www.nytimes.com/2026/01/28/travel/florida-artemis-ii-moon-launch.html
2•bookofjoe•14m ago•1 comments

Show HN: Ghostly: The Ultimate Platform for Ghosting Candidates (Satire)

https://staticfile-25978.wasmer.app/
1•dw1014•14m ago•0 comments

Neuralink has had 21 'Neuralnauts' in the past 2 years

https://neuralink.com/updates/two-years-of-telepathy/
1•ryzvonusef•15m ago•0 comments

Reasoning About Things Without Choosing Them

https://cognitivelayer.substack.com/p/chasing-the-ghosts-of-choice
1•CortexFlow•15m ago•0 comments

That's Not How Email Works, HSBC

https://danq.me/2026/01/28/hsbc-dont-understand-email/
5•HotGarbage•15m ago•0 comments

Show HN: Built an app to replace our failed group podcast experiment

https://roadsaudio.com/college
1•1manstartup•15m ago•0 comments

Native Instruments in Preliminary Insolvency

https://www.gearnews.com/native-instruments-insolvency/
3•thomas_witt•15m ago•0 comments

The new era of browsing: Putting Gemini to work in Chrome

https://blog.google/products-and-platforms/products/chrome/gemini-3-auto-browse/
5•xnx•16m ago•0 comments

Attached to Tragedy: Tracing Challenger "Remove Before Flight" Tags

https://arstechnica.com/space/2026/01/attached-to-tragedy-tracing-challenger-remove-before-flight...
1•LorenDB•17m ago•0 comments

Show HN: I built an AI companion to stop doomscrolling and regulate anxiety

https://mynomie.com/
1•liaai0630•17m ago•0 comments

Show HN: A Specialized Grok Prompt Refiner for Vision API

https://grok-bikini.com/
1•thenextechtrade•18m ago•0 comments

Students using “humanizer” programs to beat accusations of cheating with AI

https://www.nbcnews.com/tech/internet/college-students-ai-cheating-detectors-humanizers-rcna253878
2•unpredict•18m ago•0 comments

Star Trek: Planet of the Titans: The Film That Almost Changed Starfleet History

https://www.womansworld.com/entertainment/classic-tv/star-trek-planet-of-the-titans-the-epic-1976...
3•billybuckwheat•19m ago•0 comments