ClaudeDevs announces 'claude plugin eval' for Claude Code
ClaudeDevs says users can create test cases for a plugin or skill, score the runs, then rerun each case without the plugin to compare results.
TLDR
ClaudeDevs announced a new Claude Code command, 'claude plugin eval,' to help users assess what a plugin adds or whether it needs more work. The team describes a process of creating test cases, running a plugin or skill against them, scoring those runs and repeating each case without the plugin to see the differences.
Combined views
4.1K
3 Sources, first seen 19d ago