frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Mongolian Script

https://en.wikipedia.org/wiki/Mongolian_script
1•Razengan•46s ago•0 comments

The Impossible Prompt That's Easy for Humans

https://teodordyakov.github.io/the-impossible-promt/
1•ssdsa•6m ago•0 comments

Why GitHub's War on Toasts Is Bad News for Accessibility

https://medium.com/offmessageorg/why-githubs-war-on-toasts-is-bad-news-for-accessibility-a88ddbad...
2•bochoh•8m ago•0 comments

Beijing's prison silences Chinese journalist and lawyer of Covid19 in Wuhan fame

https://www.bushcenter.org/publications/the-struggle-for-freedom-beijings-revolving-prison-doors-...
1•737min•8m ago•0 comments

Show HN: MCP for finding the better NPM dependencies

https://web-production-0200a.up.railway.app
1•jsafaiyeh•12m ago•0 comments

Show HN: Native macOS timeout (Rust) – handles system sleep, no coreutils

https://github.com/denispol/darwin-timeout
1•denispol•14m ago•0 comments

Show HN: TimeProof – Timelapse your computer based workday! (MacOS)

https://apps.apple.com/us/app/timeproof/id6751671026?mt=12
1•Viper117•16m ago•1 comments

Infosys co-founder once again calls for longer than 70-hour weeks – and no, he's

https://www.techradar.com/pro/infosys-co-founder-once-again-calls-for-longer-than-70-hour-weeks-a...
4•sys_64738•17m ago•0 comments

Tesla struggles to course correct from sales skid

https://www.reuters.com/business/autos-transportation/tesla-struggles-course-correct-sales-skid-2...
3•breve•20m ago•0 comments

Show HN: Watsn.ai – Bullshit Detector

https://watsn.ai/
1•flx1012•23m ago•0 comments

Walrus: Distributed message queue written in Rust

1•peterjkl•25m ago•1 comments

How Workplace Fashion Has Changed over the Decades

https://www.wsj.com/style/fashion/office-attire-changes-history-e6e9147b
1•fortran77•26m ago•1 comments

Show HN: GitHub Activity Analytics Powered by ClickHouse

https://velocity.clickhouse.com/#org=ClickHouse&metric=all_activity&range=all&grouping=auto&alexe...
1•saisrirampur•34m ago•1 comments

AI satisfies formal structural subjectivity in the S₀ protocol

https://github.com/jengbeng/ai-structural-subject
1•jengbeng•34m ago•1 comments

Google CEO Pushes 'Vibe Coding' – But Real Developers Know It's Not Magic

https://www.interviewquery.com/p/ai-coding-vibe-coding-explained
2•birdculture•34m ago•0 comments

OBR calls in cyber expert over botched release of Budget analysis

https://www.bbc.com/news/articles/cgmn991pz9jo
1•CharlesW•35m ago•0 comments

Indus River – Scientists may have solved why this ancient civilization vanished

https://www.washingtonpost.com/climate-environment/2025/11/27/indus-river-valley-civilization-cli...
1•pseudolus•37m ago•1 comments

Manus 1.5 Uses TiDB X to Let Agents Ship Full-Stack Apps at Scale

https://www.pingcap.com/blog/how-manus-1-5-uses-tidb-x-to-let-agents-ship-full-stack-apps-at-scale/
3•ngaut•39m ago•0 comments

Is There an AI Bubble?

https://www.deeplearning.ai/the-batch/issue-329/
2•gmays•49m ago•1 comments

Gray wolf reeled in a crab trap with a rope

https://www.popsci.com/environment/wolf-uses-tool-video/
1•geox•50m ago•0 comments

Rope science, part 11 – practical syntax highlighting (2017)

https://xi-editor.io/docs/rope_science_11.html
2•todsacerdoti•51m ago•0 comments

Ask HN: White Label Chat Apps?

1•stonedge•54m ago•1 comments

A triangle whose interior angles sum to zero

https://www.johndcook.com/blog/2025/11/28/tricusp-triangle/
17•tzury•59m ago•7 comments

Virtual MCP Servers: A Use Case-Driven Solution to Tool Overload

https://www.pulsemcp.com/posts/virtual-mcp-servers-and-gateways
1•jngiam1•1h ago•0 comments

What a federal ban on THC-infused drinks and snacks could mean for hemp industry

https://apnews.com/article/hemp-thc-drinks-ban-law-8d77842cb250291a178db6f9d39f8b7e
4•petethomas•1h ago•0 comments

Show HN: Free PPT generator after NotebookLM removed the feature

https://ppt.gbase.ai
1•jinfeng79•1h ago•1 comments

Show HN: Sourcewizard – A wizard for generating integration specs

https://github.com/sourcewizard-ai/sourcewizard/tree/main/apps/planner-ui
3•mifydev•1h ago•8 comments

Air pollution may reduce health benefits of exercise

https://www.ucl.ac.uk/news/2025/nov/air-pollution-may-reduce-health-benefits-exercise
2•hhs•1h ago•0 comments

Investigating a Possible Scammer in Journalism's AI Era

https://thelocal.to/investigating-scam-journalism-ai/
2•jaredwiener•1h ago•0 comments

You Need More Lux (2023)

https://meaningness.com/sad-light-led-lux
1•turtleyacht•1h ago•0 comments
Open in hackernews

Show HN: Sourcewizard – A wizard for generating integration specs

https://github.com/sourcewizard-ai/sourcewizard/tree/main/apps/planner-ui
3•mifydev•1h ago
Hey HN, Ivan here

We just shipped the new web version of Sourcewizard — this time it’s not installing code,

planning instead the entire integration so it works on the first try.

Here's how it works: - Connect your repo - Pick any service eg: Sentry, Statsig, PostHog, Resend, Clerk - Sourcewizard deep-scans your codebase, figures out your architecture + workflows - Asks 3-6 questions about how you want to use the integration specifically in your codebase - Returns out a personalized plan: ready-to-paste prompt for Claude/Cursor

We made this because we kept fighting with cursor when doing integrations and we decided to completely open-source it.

For example you can implement Clerk auth and PostHog analytics in 10 minutes on the first attempt.

Here's our website: https://sourcewizard.ai

Twitter thread with screenshots & video: https://x.com/whysoslimm/status/1994501470398681299

The code is open source if anybody is curious: https://github.com/sourcewizard-ai/sourcewizard

Would love for HN to break it, and hope this helps someone try on a service for size quickly :)

May the force be with you: - Ivan (@ichebykin) - Lukas (@whysoslimm)

Comments

AnicetN•56m ago
i don't get all these startups these days who just wrap oai and call it a day

cool ui tho

mifydev•54m ago
please don't get confused, it's not wrapping oai at all! We're wrapping Claude Code here
AnicetN•52m ago
you mean you actually explore the code and all does it test it works?
doctorslimm•51m ago
we don't run tests for the plan in the sandbox currently but this is something we are looking at in the future
doctorslimm•51m ago
Lukas here would love to answer any questions, everyones feedback means a lot to us
RayFitzgerald•48m ago
hmm interesting. do you guys store my code post-analysis? any policies here??
mifydev•44m ago
that's one of the reasons why we open-sourced everything, we don't store the code, and you can verify that :) we're just running claude code on your codebase in an ephemeral container, only thing we save is your generated plan
animosf•43m ago
So interesting!