capa-officer

I found your capa-officer skill while reviewing some recent submissions—the concept of turning corrective action tracking into a structured tool is solid, though the execution needs some work to reach its potential at 54 points.

Links:

• GitHub
• Your agentic skill in the SkillzWave marketplace

The TL;DR

You're at 54/100, landing in F territory. This is based on Anthropic's best practices for agentic skills. Your strongest area is Spec Compliance (12/15)—the frontmatter and naming are clean. But Utility is dragging you down hard at 8/20, and PDA (Progressive Disclosure Architecture) is at 13/30. The real blocker? All your referenced files are empty placeholders.

What's Working Well

• Valid YAML frontmatter - Your metadata structure is correct with proper name conventions in hyphen-case
• Clear workflow structure - You've got numbered steps and decision points laid out logically for CAPA processes
• Good trigger terms - Description includes "CAPA investigations" and "corrective action planning" which should help discoverability
• Quality checklist bonus - You included a pre-commit checklist which added +2 points

The Big One: Empty Reference Files

Here's what's killing your score: references/api_reference.md, scripts/example.py, and assets/example_asset.txt are all empty placeholders. This breaks your entire layered architecture strategy. You've got the structure there, but no substance.

The fix: Either actually populate these with real content (investigation guides, RCA templates, decision trees) or remove the references entirely. If you add concrete examples—like showing a fishbone diagram for contamination events or a 5-Why template for single-cause issues—you'd pick up roughly +8 points easy. Right now you're telling users "see the guide" but there's no guide.

Other Things Worth Fixing

1. Add a Table of Contents - Your SKILL.md is 191 lines without a TOC. Split it into sections (Core CAPA Competencies, System Optimization, Cross-functional Integration, Resources) and add navigation links after the frontmatter. Quick +2 points.

2. Strip the marketing language - Headers like "Senior CAPA Officer" and "Expert-level" aren't instructional. Just call it "CAPA Officer" and describe what it does functionally. Same with "Expert-level Corrective and Preventive Action"—make it purely instructional, not promotional.

3. Add decision criteria - Your "Decision Point: Select appropriate RCA methodology" section is vague. Tell me when to use 5 Why (single-cause issues), when to use Fishbone (multi-factor, 3-6 factors), when to use FTA (safety-critical). Concrete decision trees here could add +3 points to Utility.

4. Show validation steps - You list workflow steps but no run-check-fix patterns. What does success look like at each stage? How do I verify the CAPA actually worked? That's the feedback loop that's missing.

Quick Wins

• Most impactful: Fill in or remove those placeholder reference files (+8 points)
• Easiest win: Add a TOC and remove marketing language (+4 points combined)
• Biggest ROI: Add one concrete example workflow with decision criteria (+5 points)

That's +17 points of low-hanging fruit that could get you to 71 territory.

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.