Content
71%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-structured standards skill with concrete, API-specific guidance and exemplary progressive disclosure: rules inline, code split into a real, one-level reference. Its main weakness is redundancy — the five core rules are restated almost verbatim in the anti-patterns section — plus the absence of an explicit writing/verification sequence.
Suggestions
Merge 'Anti-Patterns & Banned Smells' into the Core Rule Anchors (e.g., append each ban to its [MOB-TEST-0x] rule) to remove the near-verbatim duplication of all five rules and cut roughly a quarter of the body.
Remove or contextualize the empty '## **Priority: P0 (CRITICAL)**' heading, which adds no guidance as a standalone section.
Add a brief test-writing sequence (e.g., write failing reproduction per MOB-TEST-05 -> implement -> run via Gradle to verify) to raise workflow clarity with explicit checkpoints.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The five rules [MOB-TEST-01..05] are stated twice, once in "Core Rule Anchors" and again nearly verbatim in "Anti-Patterns & Banned Smells", and a stray "## **Priority: P0 (CRITICAL)**" heading adds noise, so roughly a quarter of the body is repetition. This fits 'mostly efficient but includes some unnecessary explanation or could be tightened' rather than the minor-trim level of 4. | 3 / 5 |
Actionability | The body names exact APIs ("runTest with MainDispatcherRule", "Turbine test {}", "createAndroidComposeRule + Hilt (HiltAndroidRule)", "@TestInstallIn", "composeTestRule.waitUntil") and a concrete naming example, with executable Kotlin in the referenced file. Not a 5 because there are no inline copy-paste snippets and positive examples for some bans (e.g., focused Turbine assertions) are only implied. | 4 / 5 |
Workflow Clarity | Guidance is clearly organized by test type (unit vs UI integration) with scope, DI, and a bug-first sequencing rule, and the skill is not destructive/batch so no validation cap applies. Not a 5 because there is no explicit end-to-end test-writing sequence or verification checkpoint (e.g., how to run/verify tests) beyond the implicit MOB-TEST-05 order. | 4 / 5 |
Progressive Disclosure | The body is a concise standards overview with a clearly signaled, one-level-deep References section ([Test Rules](references/implementation.md)); the referenced file exists and appropriately holds the executable code while rules stay inline, matching the 'clear overview with well-signaled one-level-deep references' anchor. | 5 / 5 |
Total | 16 / 20 Passed |