Content
35%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill provides a well-structured conceptual framework for TDD workflows but fails on actionability — it contains no executable code, no real tool commands, and relies entirely on abstract 'Task tool' delegations with undefined subagent types. It is significantly over-verbose, explaining TDD concepts and anti-patterns that Claude already knows, while the actual novel guidance (specific tool usage, concrete commands) is absent. The workflow sequencing is the strongest aspect, with clear phases and gates, though validation steps lack concrete implementation.
Suggestions
Replace abstract 'Task tool with subagent_type' steps with concrete, executable commands — actual test framework invocations (e.g., pytest commands), real code examples showing test-first patterns, and specific CLI commands for coverage reporting.
Remove sections that explain TDD concepts Claude already knows: anti-patterns, most of the Notes section, and the extended thinking block. Focus tokens on what's unique to this project's TDD setup.
Add at least one concrete end-to-end example showing a failing test, minimal implementation, and refactored result with actual code in a specific language/framework.
Split the validation checklists, metrics tracking, and configuration thresholds into a separate reference file to reduce the main skill's length and improve progressive disclosure.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is heavily verbose with extensive sections that explain TDD concepts Claude already knows (anti-patterns, notes about TDD discipline, what TDD is). The validation checklists, metrics tracking, and anti-patterns sections largely restate common knowledge. The extended thinking block and boilerplate 'Use this skill when' sections add unnecessary tokens. | 2 / 5 |
Actionability | Despite having numbered steps, the guidance is abstract and non-executable. Every step delegates to a 'Task tool with subagent_type' pattern that is not a real, documented tool — it reads like pseudocode for an orchestration system. There are zero concrete code examples, no actual test framework commands, no executable snippets, and no real CLI commands. The $ARGUMENTS placeholder is never defined. | 2 / 5 |
Workflow Clarity | The phases are clearly sequenced (Red-Green-Refactor) with explicit gates between phases and validation checklists. However, the validation steps are checklist items rather than executable verification commands. The failure recovery section exists but is vague ('rollback to last valid state' without specifying how). The workflow is well-structured conceptually but lacks concrete validation mechanisms. | 3 / 5 |
Progressive Disclosure | There is one reference to 'resources/implementation-playbook.md' but no bundle files are provided, making it unverifiable. The skill itself is a monolithic ~200-line document that could benefit from splitting detailed phase instructions, checklists, and anti-patterns into separate files. The structure has clear sections but everything is inlined. | 3 / 5 |
Total | 10 / 20 Passed |