End-to-end test-fix workflow generate test sessions with progressive layers (L0-L3), then execute iterative fix cycles until pass rate >= 95%. Combines test-fix-gen and test-cycle-execute into a unified pipeline. Triggers on "workflow:test-fix-cycle".
72
66%
Does it follow best practices?
Impact
Pending
No eval scenarios have been run
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./.codex/skills/workflow-test-fix-cycle/SKILL.mdQuality
Discovery
85%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
The description is technically detailed and specific about what the skill does and how it's triggered. Its main weakness is that the trigger term is a synthetic command rather than natural language keywords a user might use, which could reduce discoverability in natural conversation. The description is concise and well-structured with clear differentiation from related skills.
Suggestions
Add natural language trigger terms alongside the command trigger, e.g., 'Use when the user asks to run tests and fix failures, wants an automated test-fix loop, or mentions end-to-end test workflows.'
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: generate test sessions with progressive layers (L0-L3), execute iterative fix cycles, specifies a pass rate threshold (>= 95%), and names the sub-skills it combines (test-fix-gen and test-cycle-execute). | 3 / 3 |
Completeness | Clearly answers 'what' (generate test sessions with progressive layers, execute iterative fix cycles until 95% pass rate, combines two sub-skills) and 'when' (triggers on 'workflow:test-fix-cycle'). The trigger is explicit even if it's a command-style trigger rather than natural language. | 3 / 3 |
Trigger Term Quality | Includes some relevant terms like 'test-fix-cycle', 'test sessions', 'fix cycles', and 'workflow', but the trigger is a synthetic command ('workflow:test-fix-cycle') rather than natural language terms a user would say. Missing natural phrases like 'run tests', 'fix failing tests', 'test and fix'. | 2 / 3 |
Distinctiveness Conflict Risk | Clearly distinguishable with a very specific niche: end-to-end test-fix workflow with progressive layers L0-L3 and a 95% pass rate threshold. The explicit trigger command and mention of combining specific sub-skills (test-fix-gen and test-cycle-execute) make it unlikely to conflict with other skills. | 3 / 3 |
Total | 11 / 12 Passed |
Implementation
47%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill has excellent workflow clarity with well-defined phases, validation gates, error handling, and feedback loops. However, it is severely over-documented for a SKILL.md overview — agent lists are repeated 3 times, execution flows are shown in multiple formats, and content that belongs in phase docs is duplicated inline. The result is a token-heavy document that buries its actionable guidance under layers of reference material.
Suggestions
Reduce content by 50%+: Remove the ASCII architecture diagram, deduplicate agent listings (list once with phase reference), and move the directory structure, progress tracking patterns, and error handling table to a separate reference file.
Consolidate the execution flow — currently shown as ASCII diagram, text flow, and coordinator checklist. Pick one authoritative representation and reference phase docs for details.
Remove the Subagent API Reference section entirely — Claude knows how to use spawn_agent/wait_agent/close_agent, and the lifecycle pattern can be shown once in a 3-line example rather than 4 separate subsections.
Move the coordinator checklist content into the phase docs where it will be read on-demand, keeping only a brief summary in SKILL.md.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Extremely verbose at ~350+ lines. The ASCII architecture diagram, repeated agent listings (Phase 1/2 agents listed 3 times each), redundant execution flow (shown as both diagram and text), progress tracking examples, and extensive directory structures all bloat the content. Much of this (subagent API basics, what L0-L3 means conceptually) could be assumed or drastically shortened. | 1 / 3 |
Actionability | Provides concrete usage examples, subagent API code snippets, and specific commands. However, the actual execution logic is deferred to phase docs (phases/01-*.md, phases/02-*.md), making this more of a reference/overview than executable guidance. The spawn_agent/wait_agent examples are useful but the core workflow steps say 'Read phases/01-test-fix-gen.md' rather than providing the actual instructions. | 2 / 3 |
Workflow Clarity | The workflow is exceptionally well-sequenced with clear phase boundaries, a mandatory confirmation gate between phases, explicit validation checkpoints (quality gates at IMPL-001.3 and IMPL-001.5), error recovery strategies (rollback on regression, fallback chains), and iteration decision logic (<95% → fix loop, 95-99% → partial success, 100% → success). The feedback loops are explicit and well-documented. | 3 / 3 |
Progressive Disclosure | References phase docs (phases/01-*.md, phases/02-*.md) appropriately, but the SKILL.md itself contains far too much inline detail that should be in those phase docs — the full directory structure, progress tracking patterns, error handling tables, and coordinator checklists duplicate what the phase docs likely contain. The architecture diagram and task pipeline are also overly detailed for an overview document. | 2 / 3 |
Total | 8 / 12 Passed |
Validation
90%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 10 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
allowed_tools_field | 'allowed-tools' contains unusual tool name(s) | Warning |
Total | 10 / 11 Passed | |
0f8e801
Table of Contents
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.