Observed arrival · 2026-10-16
CodeTwig tests Claude Code advice against actual runs
A collection of Claude Code recipes and guides paired with reported costs, timings, diffs, and pass/fail checks.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- Developers using Claude Code on real codebases
- Worth noticing
- A subagent-search comparison used a 24-file API; delegation cost more, with no context saved.
Field notes
The homepage reports costs and timings alongside pass counts, and links to full run data for a comparison of Haiku, Sonnet, and Opus on the same feature. Its examples also distinguish task contexts: one session reportedly cost less even for unrelated tasks, while the site notes that cached context affected the result.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
✦PrettyNotable craft visible
●ProPolished or operationally mature
◎NicheUnusually specific use
One card from the complete issue