frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Show HN: Ragleap-RAG – RAG engine that documents what it doesn't do yet

https://github.com/antonyrag/ragleap-core
1•antonyragleap•2m ago•0 comments

Show HN: ArtDeck – An infinite canvas for collecting and studying references

https://getartdeck.com/
1•artdeck•3m ago•0 comments

He Wasn't Chasing Mountains

https://hasskhan.substack.com/p/he-wasnt-chasing-mountains
1•pradeepchhetri•4m ago•0 comments

Bloody Awful Work Creeps in When Standards Are Relaxed

https://dailysceptic.org/2026/08/02/bloody-awful-work-creeps-in-when-standards-are-relaxed/
2•Michelangelo11•5m ago•0 comments

Apple says MacBook Neo is pulling schools away from Windows and Chromebooks

https://9to5mac.com/2026/08/01/apple-says-macbook-neo-is-pulling-schools-away-from-windows-and-ch...
1•gslin•5m ago•0 comments

Nifty little macOS toolbar tool to track daily usage

https://github.com/shrivara/token-bar
1•Hasaber8•6m ago•0 comments

Show HN: Gainz.fast – Local Inference, Faster

https://gainz.fast/
1•carsenk•7m ago•0 comments

Show HN: Fruman – A 2D souls-like game demo running in web browser

https://fruman.baklab.app
1•Kholin•12m ago•0 comments

OpenAI has the right strategy, but is unlucky

2•Taikhoom2010•14m ago•0 comments

Trump's AI protectionism has come for robotics

https://www.technologyreview.com/2026/08/03/1141056/trumps-ai-protectionism-has-come-for-robotics/
1•joozio•14m ago•0 comments

Show HN: lde – LuaJIT written, all in one, package manager and runtime

https://lde.sh/blog/lde-0.10.0/
3•oncelock•15m ago•1 comments

Six Problems When You Try to Build an AI Therapist

https://emilylee293105.substack.com/p/six-problems-when-you-try-to-build
2•fhe•20m ago•0 comments

C64 Demo Effects Explained: Rodents in the Attic [video]

https://www.youtube.com/watch?v=uZ1atMUOUMU
1•nikanj•20m ago•0 comments

Halloween Documents

https://en.wikipedia.org/wiki/Halloween_documents
1•sundarurfriend•21m ago•0 comments

China gets better bang for its buck than America in AI

https://www.economist.com/finance-and-economics/2026/08/03/how-china-gets-better-bang-for-its-buc...
2•pingou•26m ago•0 comments

Genkan

https://en.wikipedia.org/wiki/Genkan
1•robalni•27m ago•0 comments

Three lions die at Tokyo's Tama Zoo amid heat wave

https://www.japantimes.co.jp/news/2026/08/04/japan/science-health/heat-lion-deaths-tokyo-zoo/
1•geox•30m ago•0 comments

Show HN: Visualize in 2D/3D map any GitHub user's followers

https://github.com/ThierryRakotomanana/GitCharta
1•ThierryRakt•31m ago•0 comments

ChatTJB advertises a human-powered LLM parody on an SF billboard

https://runtimewire.com/article/chattjb-human-powered-llm-san-francisco-billboard
1•ryanmerket•33m ago•0 comments

Hiring Senior SRE

https://www.axelerant.com/careers
1•adrijadas•35m ago•0 comments

I built an open-source admin panel builder for existing databases

https://github.com/Haimanot25/dboard
2•haimanot25•38m ago•0 comments

Walking to the Car Wash: The Salience Bias of LLMs in Commonsense Reasoning

https://arxiv.org/abs/2607.28478
1•theanonymousone•40m ago•0 comments

Show HN: Generate a plan visualization with dependencies to parallelize projects

https://reloadium.com/reloadium-plans/
1•julienreszka•40m ago•0 comments

Certified Program Synthesis with a Multi-Modal Verifier [pdf]

https://verse-lab.org/papers/leetproof-ase26.pdf
1•matt_d•56m ago•0 comments

Dev tools must be opensourced thats why I build this

1•Thomeras•56m ago•0 comments

Harness Engineering for Self-Improvement

https://lilianweng.github.io/posts/2026-07-04-harness/
2•tosh•58m ago•0 comments

Show HN: ePub Translator

https://epubtranslator.net/
2•xiaoape•1h ago•0 comments

Agentty ADE: preview of agentic orchestrator is out in v0.14.0

https://github.com/agentty-xyz/agentty
2•minev-dev•1h ago•1 comments

Bubo: AI code-reviewer that learns from review comments

https://github.com/mountainowl/bubo
1•mt_owl•1h ago•1 comments

An Internet for the Solar System [video]

https://media.ccc.de/v/emf2026-45-1-an-internet-for-the-solar-system
2•anticensor•1h ago•1 comments
Open in hackernews

Bubo: AI code-reviewer that learns from review comments

https://github.com/mountainowl/bubo
1•mt_owl•1h ago

Comments

mt_owl•1h ago
I'm getting tired of AI code reviewers flooding PRs with noise and repeat findings, not to say false positives. And then learning nothing when a developer explains why a finding is wrong!

For me there's a lot going on in code beyond the tech stack: tribal and institutional knowledge, business logic that evolved over time, ripple effects across upstream and downstream systems etc.

I started those requirements and added simple setup, evidence-backed findings/ LGTM, and most importantly learning from human comments on those findings, so it gets better tuned over time to each repo.

I ran a smaller benchmark, not definitive by any means, four open source reviewers on 20 pinned PRs/MRs.

  Bubo                 20/20   7/8   27 findings    0% noise
  ai-codereviewer      19/20   6/8   118 findings   20% noise
  ChatGPT-CodeReview   20/20   5/8   75 findings    11% noise
  Qodo/PR-Agent        19/20   2/8   7 findings     not scored
  Alibaba open-code-review — partial run, 4/20
All on GPT-5.5 except Qodo was on 40 (didn't support 5.5). Take it as directional results. I'll be doing more comps in coming days and upload benchmark to the repo.

I went with polling (zero steup on repo side) - wasn't trying to change the repo CI/CD's rather have a tool seamlessly work.

Roadmap item: I want to route changes to pluggable subject-matter specialists (Skills) instead of one general reviewer, i.e. industry SMEs or just an "Expert Python Guy".

Bubo is running in production in two places, a large data processing/ETL codebase and a fintech crypto stack. Would appreciate feedback, particularly on whether the learning from comments part actually holds up.