Content
55%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 workflow itself is clearly sequenced with strong validation checkpoints, error recovery, and a falsifiable verification matrix — the strongest part of the skill. However, the body is noticeably padded with repeated TDD/DDD comparisons and phase descriptions of knowledge Claude already has, the execution guidance lacks concrete commands or inline examples, and the two existing reference files are never linked from the body.
Suggestions
Cut the body to the workflow essentials: the TDD/DDD distinction is stated three times (Development Mode Configuration, Core Philosophy, Integration Points) and the phases are described twice (Implementation Guide, Standard TDD Session) — keep one authoritative copy of each and delete the rest.
Link the existing bundle files from the body, e.g. '**Examples**: See [references/examples.md](references/examples.md)' and '**Reference**: See [references/reference.md](references/reference.md)', and remove the body sections that duplicate reference.md content (phase guides, coverage table, naming conventions).
Add one concrete inline artifact to the execution steps — a minimal failing-test/example pair or the actual test-runner command (e.g. `pytest -x`) — since 'Write first failing test' and 'Run test to verify it passes' never show how to run the cycle.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Noticeably verbose with several padded sections: the TDD/DDD distinction is stated three times ('Key distinction' in Development Mode Configuration, 'TDD vs DDD Comparison' in Core Philosophy, and again in 'With DDD Workflow' under Integration Points); the RED-GREEN-REFACTOR phases are fully described in 'Implementation Guide' and then repeated step-by-step in 'Standard TDD Session' and 'TDD Loop Pattern'; and sections restate basic knowledge Claude already has ('Arrange: Set up test data and dependencies... Assert: Verify the expected outcome', 'Remove duplication / Improve naming and readability / Extract methods'). It is not anchor 1 because genuinely skill-specific material (the .moai config mode selection, coverage thresholds, TRUST validation, anti-cheat invariants, red flags) does earn its place. | 2 / 5 |
Actionability | Some concrete guidance exists ('Minimum 80% coverage per commit', '90% recommended for new code', the Verification checklist, and the two enforced anti-cheat invariants requiring verbatim RED output), but the core execution guidance is high-level — 'Write first failing test', 'Run test to verify it passes' — with no test-runner command or example test in the body, and the concrete RED-GREEN-REFACTOR code examples that exist in references/examples.md are never linked. This places it at the some-concrete-but-incomplete anchor rather than the mostly-executable anchor 4. | 3 / 5 |
Workflow Clarity | The RED-GREEN-REFACTOR sequence is explicit with per-phase validation checkpoints ('The test must fail initially', 'Run the test to confirm it passes', 'No other tests should break', 'If any test fails, revert immediately'), feedback loops for error recovery ('Refactoring Breaks Tests: Revert immediately / Refactor in smaller steps'), dedicated recovery procedures for discipline breakdown, and a final verification checklist — matching the clear-sequence-with-explicit-validation-and-recovery anchor. | 5 / 5 |
Progressive Disclosure | The body has clear section structure, but neither references/examples.md nor references/reference.md is mentioned or linked anywhere in the body — the references are present yet completely unsignaled. Additionally, references/reference.md duplicates body content (phase guidance, coverage thresholds) and ~150 lines of detailed phase/philosophy guidance that belongs in a separate file sit inline. It is above anchor 2 because the body is well-sectioned rather than a monolithic wall, but below anchor 4 because no navigation to the bundle files exists at all. | 3 / 5 |
Total | 13 / 20 Passed |