Content
51%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 well-structured as a router with excellent progressive disclosure and a concrete selection table, but it carries some over-explained testing fundamentals, lacks executable code in the body, and does not sequence an implementation workflow with validation checkpoints.
Suggestions
Trim or remove the 'Common Principles' section (AAA, Test Independence, Naming) — these are testing fundamentals Claude already applies; keep only anything repo-specific that is not in the references.
Add a short, explicit implementation workflow (e.g., resolve toolchain → write test following the selected reference → run the configured runner → verify the expected observable behavior) so the body sequences the process rather than only routing to references.
Inline one minimal concrete executable anchor (such as the runner command or a one-line import pattern) so the body has at least one copy-paste-ready action before delegating to references.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly lean, but the 'Common Principles' section re-explains Arrange-Act-Assert, test independence, and naming — basic testing concepts Claude already knows — which could be trimmed. | 3 / 5 |
Actionability | The Reference Selection table gives concrete, file-path-level routing with explicit 'When to Use' conditions, but the body itself contains no executable code or commands — all implementation detail lives one hop away in the references. | 3 / 5 |
Workflow Clarity | The body presents a routing decision and principles rather than a sequenced multi-step process; there is no explicit workflow sequence and no validation/verification checkpoints in the body. | 2 / 5 |
Progressive Disclosure | A clean overview with a well-signaled reference table pointing to two one-level-deep references (frontend.md, e2e.md, both real files), with content appropriately split between overview and detail. | 5 / 5 |
Total | 13 / 20 Passed |