vercel-plugin-eval

Verify Vercel plugin hooks by running live Claude Code sessions and analyzing debug logs to confirm hook injection and generate structured coverage data. This automated evaluation ensures reliable hook firing across runtimes and environments.

Hook dedup verification checks that repeated plugin hook injections do not duplicate actions across multiple runs. It is needed during plugin iterations to ensure regression correctness and confirm dedup logic works across environments.

Yes, you need Claude Code, WezTerm, and access to the plugin repository to run live plugin evals. These tools monitor debug logs and execute the sessions required to produce structured coverage reports.

Monitor real-time logs and generate coverage reports by launching a live Claude Code session against the plugin workflow. The session observes hook firing, collects logs, and outputs structured coverage data for analysis.

Automate end-to-end testing for Vercel plugin hook injections by running live Claude Code sessions across multiple environments. This approach validates hook firing, checks dedup correctness, and produces structured regression coverage data.