Content
67%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 dense, highly actionable reference whose tables of concrete ratios, coverage targets, and tool names are its strongest asset. Its main weaknesses are sections restating common knowledge (test file conventions, TDD basics) that inflate token cost, and a frontmatter progressive-disclosure declaration that is not backed by any bundle files.
Suggestions
Cut or compress sections Claude already knows well — the Test File Conventions table and the TDD RED-GREEN-REFACTOR quick reference — keeping only the project-specific rules ('one assertion per test (when practical)', '-race flag in Go') that are actual conventions.
Either add level-2 reference files matching the declared progressive_disclosure levels (e.g., move the pattern tables and Rationalizations into references/) or remove the unbacked progressive_disclosure frontmatter block.
Flesh out hint-style table entries like 'Interface injection, mock frameworks' and 'quickcheck, hypothesis' with one concrete example each so the patterns are directly usable rather than requiring lookup.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The tables are dense and token-efficient, but several sections restate knowledge Claude already has: the Test File Conventions table ('*_test.go', 'test_*.py', 'src/test/'), the AAA pattern explanation, and the TDD RED-GREEN-REFACTOR basics. Not 2 because most content (coverage targets with specific numbers, the Rationalizations table, Red Flags) earns its place and cannot be trimmed much further. | 3 / 5 |
Actionability | Concrete and specific throughout: exact ratios (70/20/10), numeric coverage targets (95%+, 85%+, 75%+), named tools ('go test -cover', 'istanbul', 'Stryker', 'testcontainers', 'httptest.NewServer'), and an executable Go AAA snippet. Not 5 because some table entries are high-level hints ('Interface injection, mock frameworks', 'quickcheck, hypothesis') rather than complete executable examples. | 4 / 5 |
Workflow Clarity | As a reference skill it needs no multi-step workflow, and where sequences exist they are clear (RED-GREEN-REFACTOR with explicit rules); the Verification section supplies concrete checkpoints ('run 3x to verify stability', 'show test counts per category'). Not 5 because there is no guidance on how to sequence or apply the different tables during actual test-creation work. | 4 / 5 |
Progressive Disclosure | A single well-organized file (~1,800 tokens) within the declared level2_tokens budget, with clear section headers and no buried references. Not 5 because the frontmatter declares progressive_disclosure enabled with two token levels but no level-2 bundle files exist to back that structure; not 3 since the in-file organization itself is clean. | 4 / 5 |
Total | 15 / 20 Passed |