Amplifying — AI Benchmark Research

kept by eddie feb 27

Claude Code overwhelmingly builds custom solutions over recommending third-party tools, and when it picks tools, it defaults to GitHub Actions, Stripe, and shadcn/ui.

Perhaps not Boring Technology after all
Current LLM coding agents no longer bias toward popular languages; they handle niche or new tools well via long context and documentation.
also on GitHub Actions, shadcn/ui, Vercel, Stripe, #ai-agents, #ai-coding, #software-engineering, #developer-tools
What Claude Code chooses
A report finds Opus 4.6 strongly recommends Vercel, Tailwind, shadcn/ui, and GitHub Actions, while avoiding Redis and React as default choices.
also on shadcn/ui, Vercel, Claude Code, #ai-coding, #developer-tools
Using skills to accelerate OSS maintenance
Using skills, AGENTS.md, and GitHub Actions, OpenAI accelerated OSS maintenance in their Agents SDK repos, merging 457 PRs in three months, up from 316.
also on GitHub Actions, #ai-coding, #software-engineering, #developer-tools
How Good Is AI at Coding React (Really)? - by Addy Osmani
AI coding benchmarks for React show ~40% success on isolated tasks but ~25% on multi-step integrations due to a complexity cliff, requiring strong context engineering and developer oversight.
also on shadcn/ui, #ai-coding, #software-engineering, #developer-tools
GitHub vs Vercel vs Replit: What Dev Platforms Do When AI Code Is Cheap
As AI code generation becomes cheap, GitHub, Vercel, and Replit are shifting focus to orchestration, production deployment, and verification to capture value in software development.
also on Vercel, Stripe, #ai-coding, #software-engineering, #developer-tools