Claude Code just dropped a game-changing eval command for plugins
You can now:
• Create test cases on the fly
• Score plugin performance with actual metrics
• Run A/B tests (with plugin vs without)
• See exactly where your plugin adds value or falls short
No more guessing if your custom tools actually work. This is how you iterate fast and ship quality.
Devs building on $ANTR need to check this immediately
You can now:
• Create test cases on the fly
• Score plugin performance with actual metrics
• Run A/B tests (with plugin vs without)
• See exactly where your plugin adds value or falls short
No more guessing if your custom tools actually work. This is how you iterate fast and ship quality.
Devs building on $ANTR need to check this immediately
