2. 'Coltrane' Review: Sound and Spirit (www.wsj.com)
3. Five Words from Vanishing Fleece (blog.wordnik.com)
4. MIT creates method to force AI to comply with safety rules (theframenews.org)
HardFlow, developed at MIT, forced generative AI outputs to satisfy strict safety rules in simulated robotics and imaging tests, without retraining models.
6. Understanding FlashAttention Pt 1: Personal Notes (chizkidd.github.io)
Exploring Deep Learning.
7. A list of 1,325 AI assisted repositories, mined from GitHub (github.com)
Software repository data miner. Created for my research methods masters class. - ActuallyTaylor/strata
8. Kamal – Deploy web apps anywhere from bare metal to cloud VMs (kamal-deploy.org)
From bare metal to cloud VMs using Docker, deploy web apps anywhere with zero downtime.
10. Show HN: MCP Harbor – An MCP Registry (ai.mcpharbor.dev)
11. I stress-tested Meta Muse until its agent control plane started timing out (blog.cygankiewicz.com)
Four spawn experiments against a black-box multi-agent runtime, six database lock timeouts recovered from its durable trace, and a careful look at what that evidence does and does not show.
13. Show HN: Authorize MCP tool calls without giving agents the credentials (github.com)
Keydris use-mcp tempalte for manufact with kit reader setup - keydrisLabs/mcp-auth-keydris-template
14. D-Matrix Raptor 3D-DRAM Accelerator for Generative Inference at Hot Chips 2026 (www.servethehome.com)
At Hot Chips 2026, d-Matrix showed off its Raptor 3D-DRAM accelerator for AI breaking free of using HBM for memory by stacking DRAM and logic
15. Show HN: ChaosTree – A zero-dependency Java tree library (AVL,RBT,B-Tree,B+Tree) (github.com)
Zero-dependency Java tree library featuring binary and N-ary families with JMH benchmarks and hardware-counter-backed performance evidence. - Chaos-vy/ChaosTree
16. Show HN: Bypassing Transformer Softmax via Static Contraction (github.com)
A JAX XLA-powered PoC that leverages branchless mathematical primitives to bypass the memory and execution bottlenecks of LLM softmax operations - PJHkorea/jax-softmax-bypass
17. Show HN: Pull every comment out of a Google Sheet, in the browser (bensunter.com)
Google Sheets keeps comments where you cannot sort them, count them, or hand them to anyone. Drop the workbook here and get every thread back in order.
18. Show HN: I rebuilt a 4-year-old app in 5 days using many agents – here's harness (mega.dev)
Language models got smarter, but they still lack our context. Shape the environment so agents can work without us: always on, in the cloud, with real limits.
19. Show HN: AgentDrive – persistent, versioned file storage for AI agents (tokencanopy.com)
AgentDrive is a cloud filesystem where agents and people share files across work sessions. Create multiple drives, then grant each agent or teammate access to the drives they need.
22. Show HN: Dep-radius, know which dependency updates touch your code (www.npmjs.com)
23. Show HN: Helical Time – Visualizing time as a continuous 3D torus (github.com)
visualisation of time on torus based on continuous movement of spark - nimabeh/helical-time
27. Show HN: Bare-Metal AI – Zero-dependency, sub-millisecond AI engines (github.com)
Curated list and benchmark leaderboard of ultra-lightweight, zero-dependency C/C++/Rust/Assembly engines for local LLMs, edge inference, vector search, and AI agents (2026-2027) - eminsk/awesome-ba...
28. Why the software industry needs a lot of regulation (petewarden.com)
Structural engineers have a physical stamp that they apply to plans they approve, and if a building falls down due to design problems, they’re directly and personally responsible. Software en…
29. Show HN: StemJSON – a language for LLMs to extend native mobile apps on the fly (stemjson.com)
Open declarative JSON specification for AI-generated, server-driven, end-user-customizable native mobile apps.