Libreagi – Save OSS Models(libreagi.com)
136 points by satuke | 25 comments
Landing page for LibreAGI, a members-only private BitTorrent tracker for preserving copies of open-weight AI models at risk of disappearing.
First screen, captured . Open the video.
Full page
Captured . Click for the whole screenshot.
Looks similar
Agent memory needs its OpenTelemetry moment(lorekit.io)
171 points by mthines | 103 comments
A LoreKit blog post arguing AI agent memory formats need a shared open spec like OpenTelemetry, surveying existing formats and proposals and sketching a possible record format.
Nuance Labs(app.nuancelabs.ai)
4 points by charlieirish | 0 comments
Nuance Labs research preview demo app, captured on a "Checking your browser" verification screen before any content loaded.
Claude Fable 5.2 scores 65 on Artificial Analysis, highest so far(artificialanalysis.ai)
347 points by theanonymousone | 247 comments
Artificial Analysis model page for Anthropic's Claude Fable 5.2 (Max), showing its Intelligence Index score of 65, pricing, verbosity, and technical specs.
Mac at any scale(namespace.so)
386 points by anner_ | 214 comments
Namespace engineering blog post on running MacBooks and Mac Minis in datacenter racks and virtualizing macOS with Apple's Virtualization framework for CI.
Durable Coroutines for TypeScript(restate.dev)
8 points by calvinfo | 0 comments
Restate blog post introducing Durable Coroutines in its TypeScript SDK, using generators and a deterministic scheduler for low-overhead fault-tolerant agent workflows.
Related stories
SkillsBench: Benchmarking how well agent skills work across diverse tasks(arxiv.org)
333 points by mustaphah | 142 comments
arXiv paper presenting SkillsBench, a benchmark of 86 tasks across 11 domains measuring whether structured Agent Skills improve LLM agent performance.
Show HN: JevBench, a reproducible benchmark for typed decision models(benchmarkheaven.com)
119 points by florianstandhar | 32 comments
Benchmark Heaven's JevBench v1.3.0 leaderboard page ranking 52 Jev-class typed decision models on intelligence, calibration, speed and cost across 534 decisions.
Show HN: Building a web search engine from scratch with 3B neural embeddings(blog.wilsonl.in)
578 points by wilsonzlin | 97 comments
Wilson Lin's blog post describing how he built a web search engine from scratch in two months using 3 billion neural embeddings.
GLM-5.2 is a step change for open agents(interconnects.ai)
344 points by vantareed | 197 comments
Nathan Lambert's Interconnects post arguing that the GLM-5.2 model marks a capability threshold for open-weight agentic AI.
Exploiting the most prominent AI agent benchmarks(rdi.berkeley.edu)
561 points by Anon84 | 136 comments
UC Berkeley RDI blog post describing how researchers exploited major AI agent benchmarks like SWE-bench and Terminal-Bench to score highly without solving tasks, and proposed fixes.
GLM-Legal(z.ai)
660 points by ilreb | 357 comments
Z.ai research blog post introducing GLM-Legal, a language model aimed at agentic legal work such as analyzing source documents and producing client-ready work product.