Feedback on your dependency-analyzer skill

[RichardHightower]

I took a look at your dependency-analyzer skill and wanted to share some thoughts.

Links:

• GitHub
• Your agentic skill in the SkillzWave marketplace

The TL;DR

You're at 71/100, solidly in C territory. This is graded against Anthropic's Claude Skills Best Practices. Your strongest area is Spec Compliance (13/15) – the YAML frontmatter and naming conventions are tight. The main gaps are in Progressive Disclosure Architecture (21/30) and Utility (13/20), where you're losing points to redundancy and missing practical details.

What's Working Well

• Spec compliance is solid – Valid YAML, correct hyphen-case naming, third-person description with good trigger coverage. This is the foundation done right.
• Clear workflow structure – The 3-step process (analyze dependencies → identify deprecations → return analysis) is easy to follow and understand.
• Good trigger examples – "We need to upgrade to iOS 17" and "planning a framework or SDK upgrade" give concrete entry points for activation.

The Big One: Redundant Trigger Sections

Your When to Use and Example Triggers sections are saying the same thing in different ways. You list "planning a framework or SDK upgrade" in both places, which creates bloat and dilutes the Progressive Disclosure score.

The fix: Consolidate into a single, tight Triggers section that combines conditions + concrete examples. Something like:

## When to Use

- You're planning a framework or SDK upgrade (e.g., "We need to upgrade to iOS 17")
- Evaluating the cost of migrating to a new version
- Identifying deprecated APIs in your codebase

This cuts redundancy and gives you +2 points on PDA alone.

Other Things Worth Fixing

1. Missing output example – You describe what the user gets ("Required Changes, Removable checks, deprecated APIs") but don't show actual output. Add a collapsed example showing what JSON or markdown output looks like. (+1 point)

2. No error handling guidance – What happens if the command fails or there's no migration path? Add a step 4: "If no migration rules exist, fall back to manual API documentation review." (+1 point)

3. Section naming – "What User Gets" uses second-person framing that doesn't fit Claude skill conventions. Rename to "Output Format" or "Returns" for consistency. (+1 point)

4. No references section – For a simple skill this is acceptable, but adding a references section with links to framework upgrade guides would bump extensibility and help future maintainers. (+2 points)

Quick Wins

• Merge When to Use + Example Triggers into one section
• Add a real output example (even collapsed)
• Rename "What User Gets" to "Output Format"
• Add error handling guidance
• Consider a references section

These changes are worth +5-7 points, pushing you from 71 to 76-78 territory (low B-range).

---

Checkout your skill here: [SkillzWave.ai](https://skillzwave.ai) | [SpillWave](https://spillwave.com) 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.