frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Show HN: Gnosis – Turns pull requests into guided walkthroughs

https://github.com/oddur/gnosis
3•oddurmagnusson•1h ago

Comments

oddurmagnusson•1h ago
Like many of us, I've been using Claude Code to do most of my coding and the coding part got fast, but reviewing still sucks. So either we degrade into not reviewing at all or we need to find a better way to do it. So this is my experiment to reinvent reviewing.

I've come to hate GitHub's diff view, trying to understand what it actually did. The files are shown in whatever order GitHub decides. No logical grouping. No sense of what depends on what. So with Gnosis, you paste a PR URL (or browse your open PRs), pick a model (Claude or Gemini), and it reads the diff, groups changes in a logical way, and presents them as a slideshow. Each slide explains why the change exists and shows the relevant diff. You can even comment, approve, or reject the PR directly from the tool.

It runs locally as an Electron app, to be able to use your local agent CLI tools, so it can use your existing subscriptions rather than rely on API tokens.

It´ve personally found it really useful and I would love to get feedback on how to improve it. Or even better, PRs into the repo ( guess how I will review them :)

Building an Autonomous SRE Team with AI Agents: A 5-Day Experiment

https://medium.com/@beniamin.calota/building-an-autonomous-sre-team-with-ai-agents-a-5-day-experi...
1•b3n1amin•29s ago•0 comments

Upgrading OpenClaw to Latest on Jetson Nano with Node 22

https://brtkwr.com/posts/2026-03-02-upgrading-openclaw-to-latest-node22-on-jetson-nano/
1•thunderbong•3m ago•0 comments

Qwen 3.5: small models with impressive performance

https://twitter.com/Alibaba_Qwen/status/2028460046510965160
1•moondistance•3m ago•0 comments

Show HN: OpenClaw Horror Stories – leaderboard of worst AI agent incidents

https://openclaw-horror-leaderboard.vercel.app
1•bhekanik•7m ago•1 comments

SteptronOss: Lightweight, AI-native training framework for large language models

https://github.com/stepfun-ai/SteptronOss
1•limoce•8m ago•0 comments

Ask HN: How are you structuring Markdown-based context for AI coding agents?

1•lepuski•8m ago•0 comments

Nobody Gets Promoted for Simplicity

https://terriblesoftware.org/2026/03/03/nobody-gets-promoted-for-simplicity/
1•matheusml•9m ago•0 comments

VoxCSS – CSS voxel engine for the DOM

https://voxcss.com
2•robin_reala•10m ago•0 comments

Show HN: LynxPrompt – Self-hostable, federated AI config rules manager

https://github.com/GeiserX/LynxPrompt
2•geiser•10m ago•0 comments

Algorithmic Feeds Need to Be Banned

https://shubhamjain.co/2026/02/25/algorithmic-feeds-need-to-be-banned/
1•shubhamjain•10m ago•0 comments

Addicted to the algorithm – Big Tech lobbies to keep us hooked on social media

https://corporateeurope.org/en/2026/02/addicted-algorithm-0
1•robtherobber•13m ago•0 comments

Show HN: Live Vulnerability Intelligence Dashboard (Trending CVEs)

https://www.leakycreds.com/vulnerability-intelligence
1•saynsec•13m ago•0 comments

Npmx: a fast, modern browser for the NPM registry

https://npmx.dev/blog/alpha-release
2•todsacerdoti•19m ago•1 comments

Anthropic's Killer-Robot Dispute with The Pentagon

https://www.theatlantic.com/technology/2026/03/inside-anthropics-killer-robot-dispute-with-the-pe...
2•helloplanets•20m ago•0 comments

From Abilities to AI Agents: Introducing the WordPress MCP Adapter

https://developer.wordpress.org/news/2026/02/from-abilities-to-ai-agents-introducing-the-wordpres...
1•bph•20m ago•1 comments

Show HN: CryptoMorning – AI crypto briefing via a single Claude.md file

https://plus8bit.github.io/cryptomorning/
1•plus8bit•21m ago•1 comments

.NET Devs: You're Probably Ignoring This Attribute (Stop)

https://www.youtube.com/watch?v=a7Kh8hDd96E
1•aloneguid•21m ago•0 comments

Show HN: Web2cli – Every website is a Unix command

https://github.com/jb41/web2cli
2•michaeloblak•23m ago•1 comments

India's top court angry after junior judge cites fake AI-generated orders

https://www.bbc.com/news/articles/c178zzw780xo
2•tchalla•24m ago•0 comments

Mullvad VPN: Banned TV Ad in the Streets of London [video]

https://www.youtube.com/watch?v=rwhznrpgl7k
9•vanyauhalin•28m ago•3 comments

Show HN: MRSF – A sidecar review format for Markdown (CLI and MCP server)

https://github.com/wictorwilen/MRSF
1•wictorwilen•32m ago•1 comments

OpenAI changes deal with US Military after backlash

https://www.bbc.co.uk/news/articles/c3rz1nd0egro
3•kulor•32m ago•0 comments

Show HN: Building a Globe Viewer When Software Is Cheap

https://github.com/arpentry/arpentry/blob/main/docs/MOTIVATION.md
1•bchapuis•33m ago•1 comments

Data Does Not Speak to You

https://tantaman.com/2026-03-02-data-doesnt-speak.html
1•tantaman•33m ago•0 comments

Legal AI slop is becoming a real problem

https://www.ctinsider.com/connecticut/article/supreme-court-ai-cases-middletown-21950447.php
3•dryadin•35m ago•0 comments

Show HN: Only firewall for AI prompts with a security grade on every PR

https://promptguard.co/
1•acebot712•35m ago•0 comments

Show HN: AI Referral Detector – See which AI sent traffic to your site

https://github.com/Mantasaur/AI-referral-detector
2•grandlaps•35m ago•0 comments

Startup claims it can stop lightning and prevent catastrophic wildfires

https://www.technologyreview.com/2026/03/03/1133848/this-startup-claims-it-can-stop-lightning-and...
1•joozio•35m ago•1 comments

Show HN: Claude Gym – a tiny CLI that nudges you to move while Claude Code runs

1•mosesxu•36m ago•0 comments

SDK code mode shows SotA accuracy and performance for agents using APIs

https://www.stainless.com/blog/sdk-code-mode
1•dgellow•36m ago•0 comments