Content
88%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.
A well-engineered instruction skill: it gives concrete, executable n8n guidance with real MCP tool calls, explicit decision criteria, validation checkpoints, and a closing checklist, while pushing depth into reference files. The main deductions are duplicated motivational framing and reference files that are not present in the bundle being evaluated.
Suggestions
Trim the duplicated motivation — the copy-paste-drift argument appears both in the intro paragraph and in the anti-patterns table row "Duplicating the same logic in three workflows"; one instance suffices.
Ship the referenced bundle files (SUBWORKFLOW_PATTERNS.md, NAMING_AND_DISCOVERY.md) alongside SKILL.md, or inline the small amount of essential content, so the body's pointers resolve to real files.
Condense the generic function-benefit framing ("encapsulation, reuse, testability, replaceability") to a single clause; the audience can infer standard function benefits.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dense with n8n-specific knowledge Claude does not have (Define Below vs passthrough semantics, mode all/each split, MCP tool names), but the copy-paste-drift motivation appears twice ("the same logic gets copy-pasted across workflows..." in the intro and again in the anti-patterns table) and the intro spends tokens on generic function benefits ("encapsulation, reuse, testability, replaceability"). | 4 / 5 |
Actionability | Executable MCP calls with arguments ("n8n_list_workflows()", "n8n_test_workflow({workflowId, method: \"prepare\"})"), a copy-ready trigger-node JSON snippet with typed workflowInputs, an anti-patterns table with concrete fixes, and a shipping checklist cover the common cases. | 5 / 5 |
Workflow Clarity | The extraction decision tree, the field-rename migration procedure with explicit verification ("verify with validate_workflow and n8n_get_workflow before you're done"), isolated-testing guidance, and a final pre-ship checklist give clear sequencing with explicit validation checkpoints and error-recovery direction. | 5 / 5 |
Progressive Disclosure | Good structure: deep material (in-depth mode semantics, the N+1 worked example, naming conventions) is split into two reference files surfaced via a "Read when" table one level deep. However, the referenced files (SUBWORKFLOW_PATTERNS.md, NAMING_AND_DISCOVERY.md) are not present in any bundle directory (references/, scripts/, assets/), so the navigation cannot be verified against a real bundle. | 4 / 5 |
Total | 18 / 20 Passed |