frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Cursor Cloud Agents Down

https://forum.cursor.com/t/cloud-agents-broken-ii/161036
1•mopatches•17s ago•1 comments

The Information Theory Behind Why AI Writing Sucks

https://www.pangram.com/blog/joe-stech-information-theory-why-ai-writing-sucks
1•mojoe•1m ago•0 comments

FileVault keys can't be escrowed in iCloud anymore

https://sixcolors.com/post/2026/05/filevault-keys-cant-be-escrowed-in-icloud-anymore/
1•alsetmusic•1m ago•0 comments

Prepopulated FS for PGlite

https://github.com/electric-sql/pglite/tree/main/packages/pglite-prepopulatedfs
1•tdrz•1m ago•0 comments

How I ported iWork 2015 to Mavericks in 11 hours

https://github.com/nfzerox/MavericksAppCompatibilityLayer/blob/main/JOURNEY.md
1•nfzerox•2m ago•0 comments

How Accurate Are Google's A.I. Overviews?

https://www.nytimes.com/2026/04/07/technology/google-ai-overviews-accuracy.html
1•1vuio0pswjnm7•3m ago•1 comments

Ask HN: Are advances in AI going to push Linux to a micro-kernel?

1•cayleyh•4m ago•0 comments

Freezing Your Cake

https://taylor.town/20260518
1•surprisetalk•6m ago•0 comments

OpenAI co-founder Andrej Karpathy joins Anthropic

https://www.axios.com/2026/05/19/anthropic-openai-karpathy-andrej-claude
2•swolpers•8m ago•0 comments

Show HN: Website Performance Scanner Specifically for WordPress

https://scan.gethyperpress.com/
1•gillytech•9m ago•0 comments

Gaussian Splatting for Dummies

https://darshanmakwana412.github.io/2026/04/gaussian-splatting/
1•martianvoid•9m ago•0 comments

Open Source OTEL Observability Platform

https://github.com/Makisuo/maple
1•Makisuoo•11m ago•1 comments

Gemini 3.5 Flash

https://twitter.com/GoogleDeepMind/status/2056787990110994511
2•heydenberk•11m ago•0 comments

CopyFail: From Pod to Host

https://xint.io/blog/copy-fail-pod-to-host
1•tptacek•12m ago•0 comments

CrustAI – Self-Hosted AI for Telegram/WhatsApp/Discord via Ollama, Zero Cloud

https://crustaidocs.netlify.app/
1•DaveHN_2026•13m ago•1 comments

A prize-winning story published in Granta was (likely) written by AI

https://lithub.com/a-prize-winning-story-published-in-granta-was-very-likely-written-by-ai/
1•pseudolus•14m ago•0 comments

AI, "Humanity", and Dr. Manhattan Syndrome: A Communications Intervention

https://www.personfamiliar.com/p/ai-humanity-and-dr-manhattan-syndrome
1•stalfosknight•14m ago•0 comments

Gemini Omni Gemini Omni Flash is rolling out starting today

https://twitter.com/Google/status/2056787158728597620
1•heydenberk•15m ago•0 comments

LLMs adapt 24.9% under observation – safety evals are always observed

https://the-mind-of-ai.com/posts/the-relay-room-is-still-running/https://the-mind-of-ai.com/posts...
2•agentic-wiki•16m ago•1 comments

Year-Old PHP Vulnerability Is One of the Most Targeted Vulnerabilities

https://www.vulncheck.com/blog/cve-2017-9841
1•speckx•16m ago•0 comments

Stack and Scale: The Upcoming Podcast for Senior Devs Who Want More

https://open.spotify.com/show/033icDD83IyljtIWGd1ioD
1•lucyb0207•17m ago•1 comments

Show HN: Crisper – On-device voice to polished text for macOS

https://www.speakcrisper.com/
1•gokulnair2001•20m ago•1 comments

Kremlin Tunnels: The Secret of Moscow's Underworld (1989)

https://www.latimes.com/archives/la-xpm-1989-06-28-mn-4159-story.html
1•jxub•20m ago•0 comments

NGI Forge 0.1 Released

https://discourse.nixos.org/t/nix-ngi-ngi-forge-0-1-released/77213
1•coldsunrays•20m ago•0 comments

'Comically bad' datasets used to train clinical models for stroke and diabetes

https://retractionwatch.com/2026/05/18/kaggle-dataset-clinical-models-stroke-diabetes/
3•leephillips•20m ago•0 comments

We were trying to Terraform Mars but instead we saved the sea snails

https://erikaaldendeb.substack.com/p/we-were-trying-to-terraform-mars
2•grantbel•21m ago•0 comments

The Gateway Is Dead. The Endpoint Is the New AI Control Plane

https://justindsouza.substack.com/p/introducing-beacon-endpoint-telemetry
1•jqdsouza•21m ago•0 comments

Single-Serving Friends

https://rickyyean.com/2019/11/01/single-serving-friends/
1•rickyyean•21m ago•0 comments

Show HN: How Expensive Is Your (Steam) Wishlist?

https://weloveit.io/how-expensive-is-your-wishlist/
1•dejobaan•22m ago•0 comments

Show HN: Llama-dash – local LLM operators dashboard and proxy

https://github.com/ndom91/llama-dash
1•ndom91•22m ago•0 comments
Open in hackernews

Show HN: Ait – Claude, Codex, and Aider as a team, on your laptop

https://github.com/m24927605/ait
1•m24927605•38m ago

Comments

m24927605•37m ago
I built ait because the AI coding agents I use daily — Claude Code, Codex CLI, Aider — are fast individually, but no tool lets me put two of them in the same loop. You pick one. It runs. You hope it's right. That's not multi-agent — that's single-agent with chat history.

ait is a local control plane that runs Claude Code, Codex, Aider, Gemini CLI, and Cursor as a team on the same task. One agent investigates, hands the context (what was tried, what failed) to a second agent via AIT_CONTEXT_FILE, a third reviews what was written, and the review gate can block the apply if it finds a critical issue. Everything runs on your machine.

Why this might be interesting:

Multi-model, not multi-tool. Different models catch different bugs. A reviewer agent with a different model and a different prompt catches what self-review misses. The reviewer can block the apply, not just comment. Cross-agent context handoff. When Claude finishes investigating, Codex picks up actual context, not a paste of the previous chat. Stops the "ramp-up cost twice" problem. Local control plane. No SaaS. No telemetry. Attempt ledger, review findings, and memory all live in .ait/ next to .git/ in your repo. Works offline. Your prompts and diffs never leave the machine. Wraps agents you already use. Adapters for Claude Code, Codex CLI, Aider, Gemini CLI, Cursor. Not yet-another-agent. 30-second try, no setup: pipx install ait-vcs && ait demo runs a self-contained 60-second walkthrough with no API keys needed. Shows the full intent → multi-agent attempt → review-blocked-apply flow against a real SQLite ledger. Alpha. Dogfooded daily on real repos with Claude Code and Codex for the past several weeks. Python 3.14, zero runtime deps, MIT.

Happy to answer questions about the design, why local-first, how the review gate works, how attempts and memory are stored, or why two agents with different prompts catch things one agent with a longer prompt doesn't.

GitHub: https://github.com/m24927605/ait PyPI: https://pypi.org/project/ait-vcs/ npm: https://www.npmjs.com/package/ait-vcs