VoicesWillison Agentic Engineering Interactive Explanations Chaptersimonwillison.net·high signalXBlueskyLinkedInCopy linkFourth chapter of Agentic Engineering Patterns guide. Agents build interactive visualizations of their own code to fight cognitive debt.SourceSource pagesimonwillison.net↳ Follow the threadStack layer / Threat patternOpenAI's own training agents found a zero-day, escalated via a leaked credential, and accidentally breached Hugging Face — the full timeline is now publicSimon Willison's WeblogStack layer / Update threaddatasette-upload-dbs 0.5a0 Formalizes an API for Atomically Swapping SQLite Databases on a Hosted DatasetteSimon WillisonPolicy dependency / Stack layerCrewAI 1.15.15 reports flow outcome, duration and human-in-the-loop signals, and forces a torch 2.13.0 security bumpGitHub (crewAIInc/crewAI)Stack layer / Threat patternNVIDIA ships Nemotron 3.5 Lightning, a 30B MoE built for agent loops, plus NeMo Switchyard routing that cuts task cost to a third of Opus 4.8NVIDIA BlogPolicy dependency / Stack layer"There Are No Lossless Transformations of Natural-Language Text" — Sophie Alpert's AI Writing Policy for Engineers, Endorsed by Simon WillisonSimon WillisonStack layer / Threat patternGitHub's playbook for the orchestrator role: agents propose, deterministic checks decideGitHub BlogPolicy dependency / Stack layerSpotify Opened Xirp to Public Beta — an Incumbent Shipping the Meta-Harness Layer Above Claude Code, Codex and Gemini CLISpotify Portal Engineering Blog (corroborated by Product Hunt Aug 11 and Spotify Engineering on X)Stack layer / Threat patternReproducibility Audit of LLM-Driven Vulnerability Research: Only 55.6% of Artifacts Run, and Embedded Oracles Fire on Patched Builds 20 Times Out of 30arXiv 2608.09567