Content
85%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a well-structured skill with strong actionability and workflow clarity. The examples demonstrate clear verification checkpoints after each operation, and the progressive disclosure to reference files is well-organized. The main weakness is moderate verbosity — the three examples are somewhat repetitive and some explanations (like what the check command verifies, or how to extract a presentation ID from a URL) could be trimmed since Claude can infer these.
Suggestions
Consolidate the three examples into one comprehensive example and one shorter variant to reduce repetition while maintaining coverage.
Remove the 'Cannot find presentation' troubleshooting section — Claude already understands URL structure and ID extraction.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is mostly efficient but includes some unnecessary verbosity. The setup verification section explains what the check command does (listing each dependency), the troubleshooting section explains URL structure which Claude already knows, and the three examples are somewhat repetitive — they all follow the same create-verify-insert-verify pattern. The coordinate system diagram and error handling sections are valuable and earn their tokens. | 2 / 3 |
Actionability | The skill provides fully executable bash commands with concrete arguments, specific flag names, and real-world examples. The command reference is copy-paste ready with clear argument patterns, and the examples show complete workflows with actual coordinate values and layout names. | 3 / 3 |
Workflow Clarity | Multi-step processes are clearly sequenced with explicit verification checkpoints after each major operation (create → verify with 'presentations get' → insert content → verify with 'presentations read'). Error handling distinguishes retryable from non-retryable errors with clear guidance on when to stop vs retry. The feedback loop of verify-after-each-step is consistently demonstrated across all examples. | 3 / 3 |
Progressive Disclosure | The SKILL.md provides a clear overview with well-signaled references to deeper content: permissions.md for read/write classification, command-reference.md for full argument details, layouts-guide.md, shapes-guide.md, and external OAuth/GCP setup guides. References are one level deep and clearly labeled with their purpose. | 3 / 3 |
Total | 11 / 12 Passed |