anthropic-validator

To validate Claude Code skills against Anthropic standards, run the validator on your asset file or directory. It dynamically fetches the latest documentation, identifies violations with severity ratings, and provides remediation guidance.

Yes, you can batch validate multiple Claude Code assets for compliance at once. The validator supports directory-level batch processing, producing structured YAML reports and human-readable summaries for all discovered skills, hooks, agents, commands, MCP servers, and plugins.

You can check skills, hooks, agents, commands, MCP servers, and plugins for compliance violations. The validator reviews these asset types against official Anthropic coding and structural guidelines to identify deviations.

The validator ensures assets meet the latest Anthropic coding guidelines by dynamically fetching current validation rules directly from official Anthropic documentation. This guarantees compliance checks reflect the most recent standards.

Yes, the validation report includes remediation guidance for identified compliance issues. It generates detailed reports with severity-rated findings, helping you resolve structural and coding violations before committing your assets.

Validation requires fetching up-to-date documentation dynamically, meaning an active connection is necessary to verify compliance against the latest Anthropic standards. It currently supports validating assets stored as local files or directories.