shell-testing-framework

I took a look at your shell-testing-framework skill and wanted to share some thoughts.

Links:

• GitHub
• Your agentic skill in the SkillzWave marketplace

The TL;DR

You're at 68/100, which puts you in D territory. The skill has solid bones—your Ease of Use is strong at 19/25, and your examples are genuinely helpful. But Progressive Disclosure Architecture is dragging you down hard at 12/30, and the Writing Style needs some cleanup at 5/10. This is based on Anthropic's best practices for agentic skills.

What's Working Well

• Comprehensive examples - Your test suite walkthrough for cache operations is detailed and practical. Developers can actually see how to structure tests.
• Clear metadata and triggers - "bash test", "shell script testing", "test coverage" are solid trigger terms that users will actually search for.
• Organized structure - Section headers with visual separators make it easy to navigate. The four test categories (unit, integration, edge cases, performance) are well-defined.
• Practical patterns - The arrange-act-assert pattern is explained concretely, not abstractly.

The Big One: Monolithic File Architecture

Here's what's killing your PDA score: your entire skill is crammed into a single 1335-line SKILL.md with zero reference files. This wastes tokens and violates progressive disclosure principles. Users don't need the full cache test suite example loaded every time they ask a question.

The fix: Create a references/ directory with external files:

• references/test-patterns.md - The assertion library and AAA pattern details
• references/examples/cache-tests.sh - Move the full cache test suite here
• references/performance-testing.md - Performance measurement techniques

Keep SKILL.md focused on the what and when, with brief summaries pointing to references for the how. This alone could bump you +8 points to around 76.

Other Things Worth Fixing

1. Add trigger phrases to description - Your frontmatter description lacks the "Use when asked to..." pattern. Add: "performs shell testing framework operations. Use when asked to 'shell testing', 'run bash tests', or 'test my script'." (+2 points)

2. Inconsistent voice - You mix "you must", "this is NON-NEGOTIABLE", and passive voice. Pick imperative and stick with it: "Test every function" not "Every function requires testing." Dial back the all-caps emphasis too. (+2 points)

3. Missing quick-start workflow - Add a "Quick Start: Create Your First Test File" section with 5 numbered steps. Currently you explain TDD conceptually but don't walk someone through creating an actual test file. (+3 points)

4. Examples dominate utility section - Your examples are good, but they're not templated. Add a concise "Test File Template" that people can copy-paste, then reference it rather than showing full 250-line suites inline. (+2 points)

Quick Wins

• Extract all code examples to references/examples/ folder
• Add trigger phrases to frontmatter description
• Create a numbered quick-start workflow for test creation
• Standardize voice (stick with imperative: "Do X", not "You must do X")
• Add a simple test template people can clone

These changes would realistically get you to 80-82/100. The architecture refactor is the heavy lift, but it's the thing that'll unlock the most improvement.

Checkout your skill here: SkillzWave.ai | SpillWave We have an agentic skill installer that install skills in 14+ coding agent platforms. Check out this guide on how to improve your agentic skills.