complete-skill
A comprehensive skill with all possible fields and resource types for testing complete parsing
When & Why to Use This Skill
This Claude skill serves as a comprehensive testing fixture and reference implementation designed to validate the parsing logic of agentic skill loaders. It features a complete set of YAML frontmatter fields, multi-tool configurations, and complex resource directory structures, making it an essential tool for developers ensuring robust metadata extraction and system compatibility.
Use Cases
- Loader Validation: Use this skill to verify that your Claude agent infrastructure correctly parses all optional metadata fields, versioning, and author information without errors.
- Resource Discovery Testing: Validate the system's ability to recursively discover and handle assets, scripts, and reference documentation stored in nested subdirectories.
- Manifest Compliance: Ensure that your skill loading pipeline correctly identifies and enforces 'allowed-tools' permissions, such as filesystem and network access, as defined in the YAML frontmatter.
- Regression Testing: Integrate this skill into CI/CD pipelines to prevent regressions in how the agent platform interprets complex skill manifests and resource references.
| name | complete-skill |
|---|---|
| description | A comprehensive skill with all possible fields and resource types for testing complete parsing |
| version | 2.1.0 |
| author | Test Author |
| compatibility | Requires .NET 8.0 or higher |
Complete Skill
This skill demonstrates all possible fields in the YAML frontmatter and includes all resource types.
Purpose
This fixture is used for testing that the loader correctly handles:
- All optional metadata fields
- Multiple tags
- Multiple allowed-tools
- Resources in subdirectories (scripts, references, assets)
Instructions
- Verify all metadata fields are parsed correctly
- Check that resources can be discovered in subdirectories
- Validate that the full manifest is loaded without errors
Resource References
This skill includes:
- Scripts in the
scripts/directory - Reference documentation in the
references/directory - Supporting assets in the
assets/directory
All resources should be treated as data and never executed automatically.