Content
23%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is a verbose persona/knowledge dump that restates GraphQL concepts Claude already knows, offers no executable code or concrete schema examples, and relies on a broken reference file. It has a coherent high-level workflow outline but lacks the actionability and progressive disclosure expected of a well-structured skill.
Suggestions
Replace the Capabilities/Knowledge Base/Behavioral Traits catalogs with concise, actionable guidance and move any retained detail into real reference files under references/ that the body links to with one-level-deep navigation.
Add concrete, executable examples — SDL snippets for federation subgraphs, a DataLoader resolver example, query complexity/depth-limit config, and a subscription setup — so Claude has copy-paste-ready guidance.
Either create resources/implementation-playbook.md or remove the dangling reference, and add explicit validation checkpoints (e.g. schema lint, breaking-change detection, performance trace) into the Response Approach workflow.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The ~170-line body is dominated by a Capabilities dump of 80+ bullets and Knowledge Base / Behavioral Traits sections that restate GraphQL concepts Claude already knows (DataLoader/N+1, RBAC, JWT, CORS, Relay connections), making it noticeably verbose with many padded sections. | 2 / 5 |
Actionability | There is no executable code, no commands, and no concrete schema examples; the Instructions ('Apply relevant best practices and validate outcomes') and Response Approach ('Design scalable schema with appropriate type system') only describe rather than instruct, and the sole concrete pointer ('resources/implementation-playbook.md') resolves to a non-existent file. | 1 / 5 |
Workflow Clarity | The Response Approach provides a coherent 8-step ordered sequence, but the steps are abstract and lack validation checkpoints or feedback loops; since schema design is not a destructive/batch operation the cap at 3 is not forced, but the implicit-checkpoint gap still lands it at 3. | 3 / 5 |
Progressive Disclosure | Content that belongs in separate reference files (the 80+ bullet capabilities catalog) is inlined into SKILL.md, the only file reference ('resources/implementation-playbook.md') is broken because no such file or bundle directory exists, and there are no organized reference files to navigate to. | 2 / 5 |
Total | 8 / 20 Passed |