Claude Plugin Validator FAQs

Question 1

What are the severity levels in the validation report?

Accepted Answer

Results are categorized as CRITICAL (errors that prevent loading), WARNING (potential malfunctions or deprecated fields), and INFO (best practice suggestions and optimization opportunities).

Question 2

Does it support validation of different Claude models?

Accepted Answer

Yes, it specifically checks frontmatter to ensure specified models are valid options such as Sonnet, Opus, or Haiku, or set to inherit correctly.

Question 3

How do I trigger a health check for my plugin?

Accepted Answer

You can use natural language triggers like 'validate plugin', 'check plugin health', or 'why is my plugin not loading' to start the validation process within Claude Code.

Question 4

What does the Plugin Validator actually check?

Accepted Answer

It validates the plugin.json manifest for correct JSON syntax, required fields, and semantic versioning. It also checks that all referenced agent and skill files exist, validates YAML frontmatter in those files, and detects circular dependencies.

Question 5

Can it detect broken file references?

Accepted Answer

Yes, it maps the entire directory structure against the manifest to ensure every agent, command, and skill path is valid and uses relative file paths.

Claude Plugin Validator

주요 기능

사용 사례

Claude Plugin Validator

주요 기능

사용 사례