Lance Martin to add data-first walkthrough to Claude eval skill after Hamel Husain feedback
RLanceMartin · x · 2026-10-01
Responding to review feedback from Hamel Husain and Isaac Flath, Lance Martin says his latest Claude skill—which instructs Claude to build a viewer for eval examples—lacks a step walking users through the data first. He agrees this helps prioritize which evals to write and will add it, a small but telling iteration in eval-engineering skill design.
More from coding & agent
- HF researcher admits a well-maintained monorepo is the superior way to run an AI lab — soldni · 2026-10-01
- uv replaces five Python tools at 10-100x pip speed, written in Rust — blaizedsouza · 2026-10-01
- KV-streams trains SWE agents 2x faster by preserving KV cache across compaction — burny_tech · 2026-10-01
- LangChain uses 'human touches per PR' as a key metric to evaluate coding agents — LangChain · 2026-10-01
- LangChain dev shows Slack-triggered coding agent that auto-addresses PR comments — LangChain · 2026-10-01
- CorpusMap organizes growing corpora by recurring entities so agents follow links and save tokens — coallaoh · 2026-10-01