frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Nearly a million passports and photo IDs were unprotected on the public internet

https://www.theverge.com/tech/947157/passports-data-breach-cannabis-club-systems-nefos-puffpal
1•N_A_T_E•16s ago•0 comments

Show HN: Rapunzel – a tree-style tab terminal emulator for Codex Claude Gemini

https://github.com/javaid-codes/rapunzel/tree/main
1•WasimBhai•54s ago•0 comments

My one weird trick for managing my internet addiction

https://scottrogowski.com/my-one-weird-trick-for-managing-my-internet-addiction
1•scottrogowski•1m ago•0 comments

Brendan Sorsby, Texas Tech part ways after gambling scandal, legal fallout

https://www.on3.com/news/brendan-sorsby-texas-tech-mutually-part-ways-after-gambling-scandal-lega...
1•randycupertino•2m ago•0 comments

" Claude Fable 5 was delisted. The answer isn't a bigger model – it's a panel: "

https://www.orcarouter.ai/blog/model-fusion-in-production-inside-orcarouter-fusion-and-the-routin...
1•sangwen•4m ago•0 comments

Snap Keynote at AWE USA 2026

https://www.youtube.com/watch?v=bsqQG4iiK2Y
2•kensai•7m ago•2 comments

He Won't Stop Building a Map to an Imaginary Place [video]

https://www.youtube.com/watch?v=Is8N7B9b0GQ
1•Fraterkes•8m ago•0 comments

Rare medieval bookmark exceeds expectations at auction

https://www.thehistoryblog.com/archives/76314
1•speckx•8m ago•0 comments

AI and brain-computer interface allow speechless ALS patient to work full-time

https://www.theregister.com/science/2026/06/16/ai-and-brain-computer-interface-allow-speechless-a...
3•geekinchief•9m ago•1 comments

State of Tech Hiring 2026: What 106,000 Job Listings Reveal

https://www.jobscroller.net/state-of-tech-hiring-2026
1•couentine•9m ago•0 comments

The Anthropic Fable saga proves: we have opened the AI Pandora's box. What now?

https://www.theguardian.com/commentisfree/2026/jun/16/anthropic-fable-ai
1•kuerbel•9m ago•0 comments

Show HN: Azure DevOps TUI Management Style

https://github.com/elpulgo/azdo
1•elpulgo•10m ago•0 comments

I took the le chaton fat AI model meme a step further by making a model

https://github.com/princezuda/lechatonfat
2•zuda•10m ago•1 comments

Polymarket Traders Clash over $345M Iran Peace Market (2026)

https://www.bloomberg.com/news/articles/2026-06-15/polymarket-traders-clash-over-345-million-iran...
1•kratom_sandwich•11m ago•0 comments

Y2Social – a Y2K-era social network I've been building solo

https://www.y2social.com
1•Jason463•12m ago•0 comments

Bug in FIFA World Cup internal system gave anyone ability to modify TV stream

https://techcrunch.com/2026/06/16/bug-in-fifa-world-cup-internal-system-gave-anyone-ability-to-mo...
1•cdrnsf•13m ago•1 comments

Future Self

https://futureyou.media.mit.edu/
2•onemoresoop•14m ago•0 comments

Show HN: Almost all of MonsterWriter's back end is open source

https://www.monsterwriter.com/building-linkedrecords.html
1•WolfOliver•14m ago•0 comments

Apple is about to make Hide My Email useless

https://arseniyshestakov.com/2026/06/16/apple-is-about-to-make-hide-my-email-useless/
3•SXX•17m ago•0 comments

Ask HN: What's your multi-agent orchestration setup, and success rate with it?

1•mandeepj•17m ago•1 comments

A Compendium of Canonical Charts

https://chris-parmer.com/compendium-of-canonical-charts/
1•chriddyp•24m ago•0 comments

Users cry foul after AMD stripped memory crypto from its consumer CPUs

https://arstechnica.com/security/2026/06/users-cry-foul-after-amd-stripped-memory-crypto-from-its...
9•Cider9986•24m ago•0 comments

Ask HN: How do you make LLM generated text believable?

1•StahlGuo•24m ago•3 comments

Ask HN: Those making $500/month on side projects in 2026 – Show and tell

2•fazkan•25m ago•1 comments

Why Anthropic candidates fail culture after clearing coding and system design

https://www.hack2hire.com/blog/what-anthropic-actually-tests-and-what-gets-candidates-rejected-2026
2•hack2hire•25m ago•2 comments

Run Gemma on the edge with the Coral Board [video]

https://www.youtube.com/watch?v=o2rUT2GloV0
1•simonpure•25m ago•0 comments

US manufacturer of military TV walls sold to China, US wants it back

https://www.thewirechina.com/2026/06/14/behind-the-screens/
3•ilamont•26m ago•0 comments

German broadcaster removes TV intro after Elon Musk takes legal action

https://www.bbc.com/news/articles/c0jyzp9z9deo
1•only_in_america•28m ago•0 comments

Will AI End the Open Internet? [Wading Through AI – Episode 6] [video]

https://www.youtube.com/watch?v=gR2T1uxHG7o
1•justin66•29m ago•0 comments

A 1969 camera operators' strike created Upstairs Downstairs multiverse

https://ironicsans.ghost.io/the-color-strike/
2•ohjeez•30m ago•0 comments
Open in hackernews

Show HN: VulnFeed – 9 security tools your AI agent can call (MCP server)

https://vulnfeed.novadyne.ai/
1•ngburke•1h ago

Comments

ngburke•1h ago
I built this because I kept asking Claude to check my dependencies for known vulns and it would either make up CVEs or tell me to go run something myself. VulnFeed is an MCP server that gives an agent real, current vulnerability data it can actually call.

The useful bit in one line: point it at a `package-lock.json` / `requirements.txt` / `go.sum` / etc. and it returns the known CVEs, each scored by EPSS (the probability it'll actually be exploited — so you triage the 2 that matter instead of the 40 that don't), plus the fixed version.

Try it in 10 seconds, no signup, no key:

    curl -s -X POST https://vulnfeed-api.novadyne.ai/vulnscan/query \
      -H "Content-Type: application/json" \
      -d '{"packages":[{"name":"lodash","version":"4.17.4","ecosystem":"npm"}]}'
Or wire it into Claude / any MCP client:

    {"mcpServers":{"vulnfeed":{"command":"uvx","args":["vulnfeed-mcp"]}}}
then ask it to "scan my package-lock.json."

Boring details: the client is MIT (github.com/novadyne-hq/vulnfeed-mcp, on PyPI as `vulnfeed-mcp`). Scanning is free. Two of the nine tools — detailed single-CVE lookup and continuous project monitoring — are paid; you can pay per call with no account at all (HTTP 402 micropayment) or subscribe, so there's no signup wall just to try it. The hosted backend is closed for now.

I'd genuinely like feedback on three things: (1) what tools are missing that you'd want your agent to have, (2) does per-call vs subscription pricing make sense here, (3) how's the data freshness / any false positives you hit.