fp.

Cancer blood tests are everywhere. Do they work?

https://www.nature.com/articles/d41586-026-00661-2

1•bookofjoe•58s ago•0 comments

Congress puts the ISS on life support until 2032, orders Moon base plan

https://www.theregister.com/2026/03/05/iss_extension/

1•voxadam•1m ago•0 comments

Show HN: A Linter for Skill.md Files

https://github.com/tilework-tech/nori-lint

1•theahura•1m ago•0 comments

Wikipedia in read-only mode following mass admin account compromise

https://www.wikimediastatus.net

2•greyface-•1m ago•1 comments

Show HN: hugpoint.io – Find the fairest meeting spot for a group

https://hugpoint.io/

1•prunax•3m ago•0 comments

Show HN: ExtraSuite – pull / push for Google Drive

https://github.com/think41/extrasuite

1•ksri•4m ago•0 comments

FCC Chair to Europe: If You Restrict US Satellite Providers, We'll Ban You Here

https://www.pcmag.com/news/fcc-chair-to-europe-restrict-us-satellite-providers-well-ban-you-in-us

1•msolujic•4m ago•0 comments

Bumper-Sticker Computer Science

http://quotes.cat-v.org/programming/bumper-sticker-computer-science

1•Spide_r•4m ago•0 comments

Show HN: I built a simple meditation app because I was tired of subscriptions

https://apps.apple.com/us/app/pure-meditate/id6759008961

1•PureMeditate•5m ago•0 comments

Show HN: Make beats, produce music from the command line

3•lmohseni•5m ago•0 comments

The Pentagon-Anthropic feud is quietly obscuring the real fight over military AI

https://www.fastcompany.com/91502340/the-pentagon-anthropic-feud-is-quietly-obscuring-the-real-fi...

2•johnshades•5m ago•0 comments

Show HN: Reduce LLM token use by ~30% with this MCP/CLI tool(Claude benchmarked)

2•jahala•6m ago•1 comments

Elo-Driven Development

https://faizank.substack.com/p/elo-driven-development

1•fazkan•7m ago•0 comments

Kristi Noem Misled Congress About Top Aide's Role in DHS Contracts

https://www.propublica.org/article/kristi-noem-dhs-misled-senate-judiciary-corey-lewandowski-cont...

1•SilverElfin•7m ago•1 comments

Evervault raises $25M to build the clearinghouse for sensitive data

https://evervault.com/blog/series-b

2•ShaneCurran•8m ago•0 comments

Re-exposure to reward re-evaluates related memories

https://www.sciencedirect.com/science/article/pii/S0960982225015969

1•PaulHoule•8m ago•0 comments

HATEOAS Works with an LLM in the Mix

1•charlieflowers•9m ago•0 comments

Show HN: Glad-IA-Tor – Roast vibecoded products trying to get some Reach

https://glad-ia-tor.com/fr

2•GiornoJojo•9m ago•1 comments

"Too Clever by Half" (Thanks HN)

https://owl.billpg.com/too-clever-by-half/

1•billpg•9m ago•0 comments

Agentic Code Reasoning

https://arxiv.org/abs/2603.01896

1•gmays•9m ago•0 comments

LexisNexis confirms data breach, some customer records affected

https://www.theregister.com/2026/03/04/lexisnexis_legal_professional_confirms_data/

2•johnshades•10m ago•0 comments

Show HN: Webmcp-react – React hooks that turn your website into an MCP server

https://github.com/mcpcat/webmcp-react

3•kashishhora•10m ago•0 comments

Show HN: Pre-execution verification for LLM-generated agentic workflows

https://github.com/le0li0n/workflow-verify

2•jaredwaxman•11m ago•1 comments

Defensive Refusal Bias in LLMs Is Hurting Infosec

http://lockboxx.blogspot.com/2026/03/defensive-refusal-bias-in-llms-is.html

2•ahokk•11m ago•0 comments

Show HN: OpenRouter Skill – Reusable integration for AI agents using OpenRouter

https://github.com/bnishit/openrouter-skill

2•bnishit•12m ago•0 comments

Show HN: KeepFiled – forward an email and it automatically files in Google Drive

https://www.keepfiled.com/

2•popupeyecare•12m ago•0 comments

Show HN: Trueline – Hash-verified edits save 44% of Claude's output tokens

https://github.com/rjkaes/trueline-mcp

2•rjkaes•14m ago•0 comments

Apple: Enough Is Enough

https://bastibe.de/2026-03-05-apple-woes.html

3•speckx•15m ago•0 comments

Collective hum of black holes could mend our broken understanding of universe

https://www.livescience.com/physics-mathematics/collective-hum-of-black-holes-could-mend-our-brok...

1•Brajeshwar•15m ago•0 comments

A once-proud tradition is becoming awkward for elite universities

https://www.economist.com/united-states/2026/03/04/a-once-proud-tradition-is-becoming-awkward-for...

2•shadow28•16m ago•1 comments

Cancer blood tests are everywhere. Do they work?

Congress puts the ISS on life support until 2032, orders Moon base plan

Show HN: A Linter for Skill.md Files

Wikipedia in read-only mode following mass admin account compromise

Show HN: hugpoint.io – Find the fairest meeting spot for a group

Show HN: ExtraSuite – pull / push for Google Drive

FCC Chair to Europe: If You Restrict US Satellite Providers, We'll Ban You Here

Bumper-Sticker Computer Science

Show HN: I built a simple meditation app because I was tired of subscriptions

Show HN: Make beats, produce music from the command line

The Pentagon-Anthropic feud is quietly obscuring the real fight over military AI

Show HN: Reduce LLM token use by ~30% with this MCP/CLI tool(Claude benchmarked)

Elo-Driven Development

Kristi Noem Misled Congress About Top Aide's Role in DHS Contracts

Evervault raises $25M to build the clearinghouse for sensitive data

Re-exposure to reward re-evaluates related memories

HATEOAS Works with an LLM in the Mix

Show HN: Glad-IA-Tor – Roast vibecoded products trying to get some Reach

"Too Clever by Half" (Thanks HN)

Agentic Code Reasoning

LexisNexis confirms data breach, some customer records affected

Show HN: Webmcp-react – React hooks that turn your website into an MCP server

Show HN: Pre-execution verification for LLM-generated agentic workflows

Defensive Refusal Bias in LLMs Is Hurting Infosec

Show HN: OpenRouter Skill – Reusable integration for AI agents using OpenRouter

Show HN: KeepFiled – forward an email and it automatically files in Google Drive

Show HN: Trueline – Hash-verified edits save 44% of Claude's output tokens

Apple: Enough Is Enough

Collective hum of black holes could mend our broken understanding of universe

A once-proud tradition is becoming awkward for elite universities

How prompt caching works in Claude Code: experiments and architectural lessons

Cancer blood tests are everywhere. Do they work?

Congress puts the ISS on life support until 2032, orders Moon base plan

Show HN: A Linter for Skill.md Files

Wikipedia in read-only mode following mass admin account compromise

Show HN: hugpoint.io – Find the fairest meeting spot for a group

Show HN: ExtraSuite – pull / push for Google Drive

FCC Chair to Europe: If You Restrict US Satellite Providers, We'll Ban You Here

Bumper-Sticker Computer Science

Show HN: I built a simple meditation app because I was tired of subscriptions

Show HN: Make beats, produce music from the command line

The Pentagon-Anthropic feud is quietly obscuring the real fight over military AI

Show HN: Reduce LLM token use by ~30% with this MCP/CLI tool(Claude benchmarked)

Elo-Driven Development

Kristi Noem Misled Congress About Top Aide's Role in DHS Contracts

Evervault raises $25M to build the clearinghouse for sensitive data

Re-exposure to reward re-evaluates related memories

HATEOAS Works with an LLM in the Mix

Show HN: Glad-IA-Tor – Roast vibecoded products trying to get some Reach

"Too Clever by Half" (Thanks HN)

Agentic Code Reasoning

LexisNexis confirms data breach, some customer records affected

Show HN: Webmcp-react – React hooks that turn your website into an MCP server

Show HN: Pre-execution verification for LLM-generated agentic workflows

Defensive Refusal Bias in LLMs Is Hurting Infosec

Show HN: OpenRouter Skill – Reusable integration for AI agents using OpenRouter

Show HN: KeepFiled – forward an email and it automatically files in Google Drive

Show HN: Trueline – Hash-verified edits save 44% of Claude's output tokens

Apple: Enough Is Enough

Collective hum of black holes could mend our broken understanding of universe

A once-proud tradition is becoming awkward for elite universities