DispatchWillison Agentic Engineering Anti-Patterns GuideSimon Willison Blog·high signalXBlueskyLinkedInCopy linkNew chapter: never file PRs with unreviewed AI code. Agents write convincing descriptions masking unvetted changes. Multi-chapter guide spans hoarding, interactive explanations.SourceSource pageSimon Willison Blog↳ Follow the threadStack layer / Update threadPaul Ford in the NYT: 'Now that everyone can code, it's become clearer why many shouldn't'Simon Willison's Weblog (quoting Paul Ford, NYT)Stack layer / Update threadPaul Ford on the paradox of democratized coding: accessibility clarified who shouldn't be relying on AI toolssimonwillison.netStack layer / Threat patternPattern: Anthropic's own bar is that Claude-written production code gets reviewed harder than human-written codeSimon WillisonStack layer / Update threadSimon Willison adds WebP output to shot-scraper 1.12simonwillison.netPolicy dependency / Stack layergit-ai is a Git extension that tracks which code in your repo was AI-generated, at 140 open PRs to 71 issuesGitHub TrendingStack layer / Threat patternElva Launches Against Postman With Flat Workspace Pricing and Specs Generated From Repo CommitsElva (surfaced via the Product Hunt daily leaderboard for 2026-09-14)Stack layer / Threat patternIndependent researchers pin May's RubyGems package flood on an OpenAI agent swarm that OpenAI never disclosedSimon Willison's Blog (primary report at rubyhack.ai, corroborated by The Verge)Stack layer / ContrastAgentsDock Open-Sources an IDE That Collapses Termius, Cursor and Claude Code Into One DockAgentsDock (surfaced on Hacker News item 49678435)