A curated collection of noteworthy tools, articles, and resources from across the web — searchable, organized, and enriched with AI analysis.
Showing 97-120 of 1252 bookmarks
All Stories
Apple’s official Agent Skills exported from Xcode 27 — SwiftUI, UIKit modernization, Swift Testing, C bounds-safety, and security hardening for AI coding agents. - superagents-lab/xcode27-skills
Serve Qwen3.8-Flash-Next (turboderp ExLlamaV3 pack) on one NVIDIA DGX Spark with vLLM + vllm-exl3 - vcruz305/Qwen3.8-Flash-Next-EXL3-DGX-Spark-recipe
Contribute to NandhaKishorM/laya development by creating an account on GitHub.
Run typed option logits and autoregressive JSON locally with open models and WebGPU.
Bonsai Demo. Contribute to PrismML-Eng/Bonsai-demo development by creating an account on GitHub.
Ternary Bonsai 2 27B retains 98.2% of Qwen3.8 27B benchmark performance in a 5.9GB footprint, with multimodal and agentic capabilities.
Let AI agents use your real, logged-in browser without interrupting your work. CLI + extension for browser automation across any shell-capable AI agent. - Tencent/BrowserSkill
Descript is the AI video and audio editor that makes editing as easy as editing text. Record, transcribe, edit, and publish in one tool. Try it free.
Training metrics of the mimo-v2.6-pro and mimo-v2.6-flash reinforcement-learning runs, live from the trainer’s logs.
In September 2026, NVIDIA announced it is leaning into native GPU programming in Rust. CUDA C++ and CUDA Python are mature, enterprise-grade toolchains, and NVIDIA will be growing and maturing CUDA…
...or how to make Qwen learn query optimization via agentic reinforcement learning
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
A useful pattern for LLM classification at scale is to let it hallucinate plausible, fake entities. Then resolve to real ones.
System One launched Jev, a model restricted to structured typed output for tasks like classification, routing, and scoring, rather than general-purpose text generation. It uses RLCD to emit calibrated confidence values per result, which the company labels "can't hallucinate." HN commenters challenged that framing, noting the model can still produce valid but incorrect values and that constrained LLMs already enforce structured output.
TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access.
A portable agent skill for designing MCP tools, diagnosing integrations, and reviewing complete tool contracts. - RuggG/mcp-engineering
Why retro currently funnels every reviewer-agent gap into the same fix: add a line to CODING_STANDARDS.md. That's the right call for a judgement call, but for a fixed syntactic pattern it just…
Our CI job volume increased 25x over 6 months. We patched our test selection service three times before finding a sustainable solution.
How Anthropic runs AI incident response for CI/CD: an engineer on our Continuous Integration team walks through the Claude Tag agent that detects, triages, and resolves CI failures.
The Hacker News thread compares Apple's Safari MCP with Chrome DevTools MCP and Playwright MCP, noting DevTools MCP offers deeper access to performance profiling, network requests, and Lighthouse metrics. Users debate testing Safari without Apple hardware, with critics citing Apple's walled-garden strategy and defenders calling real-device testing standard across all platforms. The Safari MCP excludes personal data like AutoFill and login sessions, and participants raised concerns about constraining agent actions within a logged-in browser.
Native Safari browser automation for AI agents. 97 tools via AppleScript — zero overhead, keeps logins, runs silently in background. Drop-in alternative to Chrome DevTools MCP with 40-60% less CPU/…
Makefaster.dev is an AI skill that runs an autoresearch loop to continuously discover, test, and implement performance improvements.
qwen38-flash-next recipe for solo GB10. Contribute to bilikaz/qwen38-flash-next-recipe development by creating an account on GitHub.
Experimental exploration of skills discovery and distribution through MCP primitives. Maintained by the Skills Over MCP Working Group. - modelcontextprotocol/ext-skills
