Weekly Best
3061.
3062.
Smallweb Is Becoming an Archipelago
(hypersubject.net)
3063.
Real-time AI voice bridges language gaps in the moment
(manateavagner.com)
3064.
Open Source Agent, Harness-1, Outperforms GPT-5.4 on Recall
(venturebeat.com)
3065.
Ramp Applied AI Solutions
(ramp.com)
3066.
3067.
3068.
3069.
Port React compiler to Rust (merged)
(github.com)
3070.
Show HN: Persist.chat – Outreach and Sales Agents
(persist.chat)
3071.
3072.
Factoring "short-sleeve" RSA keys with polynomials
(blog.trailofbits.com)
3073.
I'm simulating the 2026 World Cup with 22 LLM-written agents per match
(agentpitch.surge.sh)
3074.
3075.
3076.
The Dashlane 2FA Breach and What It Means for Cloud Vaults
(cloudlesssoftware.com)
3077.
Rationales for Standards
(thomas-huehn.com)
3078.
3079.
Gordon Wood, RIP
(nytimes.com)
3080.
AWS App Runner availability change
(docs.aws.amazon.com)
3081.
Show HN: Email and identity stack for AI Agents
(mailgent.dev)
3082.
Are high-end knife shops worth markup above big cutlery chains?
(paragon-knives.com)
3083.
A little progress is worth a trillion dollars
(abundanceandgrowth.org)
3085.
3086.
3087.
3088.
New experimental versions of TeXmacs for Windows 11 and Android (revision 15512)
(lists.texmacs.org)
3090.
Claude Fable 5 and new AI safety fables
(interconnects.ai)