Data-driven studies of AI coding assistants (Claude Opus 4.7, Codex, Cursor, etc.)
31 articleWhen Opus 4.7 builds anything non-trivial, it reaches for a pnpm workspace. 454 monorepos and counting.
A first look at what Claude's flagship coding model builds when unleashed on GitHub.
Behind the scenes of the Repobility corpus — harvesting, analysis, embedding, and distillation.
The full language footprint of 9,281 Opus 4.7 repos — top 15 primary languages and the file-level extension distribution.
Hand-graded top picks from 9,281 Opus 4.7 repos — the best MCP servers, CLIs, web apps, and libraries.
18% of Opus 4.7 repos contain a CLAUDE.md. Here's what's in them and why it matters.
Despite the frontend hype, nearly half of Opus 4.7's GitHub output is backend code.
Our research is based on continuous analysis of 128,000+ repositories and 3.27 billion lines of code using Repobility's proprietary scanning engine.
All data is aggregated and anonymized. No individual repository names or source code is disclosed.
Access our proprietary datasets for your own research, product development, or competitive intelligence.
Browse DatasetsGet our latest research and intelligence reports delivered to your inbox.
No spam. Unsubscribe anytime.