Jane Street's Yinsky shares tips for making codebases agent-friendly: types, property tests, formal methods
sh_reya · x · 2026-10-27
A talk by Jane Street's Yaron Minsky on adapting codebases for AI agents is circulating widely. Key points: types are among the best feedback signals for agents — make illegal states unrepresentable and global rules get enforced across the codebase; use property-based tests, fuzzing, and full-system deterministic simulation testing; adopt expect tests that print and capture behavior to shrink review surface; and formal methods are becoming viable as agents change the ROI — seL4's 8,700 lines of C took 25 person-years with 23 lines of proof per line of code.
More from coding & agent
- Practical tip: use max effort for planning, xhigh for implementation in Claude Code — majidmanzarpour · 2026-10-27
- Chrome team explains WebMCP: agents get tools from your website, not a replacement for MCP — rseroter · 2026-10-27
- Agent-Native apps: why agents bypassing your UI can break business rules — rseroter · 2026-10-27
- 'My intelligence stack': developer splits AI coding work across Opus 5.5, Fable and Astra with human review gates — sh_reya · 2026-10-27
- Baseten webinar: evaluate open models for coding agents by benchmarking on your own merged PRs — baseten · 2026-10-27
- Thorsten Ball on working with agents: run more evals, automate everything, use the best models — omarsar0 · 2026-10-27