advocate-sourcing

I've been digging through the skill grading results and noticed your advocate-sourcing skill landed at 62/100 — the scoring suggests some solid foundational thinking around sourcing strategies, but there's clearly room to tighten up the documentation and make the use cases more concrete for developers actually trying to implement this.

Links:

• GitHub
• Your agentic skill in the SkillzWave marketplace

The TL;DR

You're at 62/100, which puts you in D territory. This is based on Anthropic's 5-pillar rubric for skill quality. Your strongest area is Spec Compliance (11/15) — the frontmatter is solid and naming is correct. The weak spot is Progressive Disclosure Architecture (16/30) — everything's crammed into one file with no reference structure to organize the complexity.

What's Working Well

• Valid frontmatter structure — Your YAML is clean and follows conventions properly
• Clear framework thinking — The 5-step workflow (Signal Collection → Scoring → Tiering → Backlog Management → Consent Tracking) is logically sound
• Hyphen-case naming — advocate-sourcing follows the right conventions

The Big One: Missing PDA Structure (High Priority)

Right now, you've got everything in a single SKILL.md file. The report notes you have zero reference files, which kills your PDA score (-6 points potential gain). Here's the fix:

Create a references/ directory with three files:

• references/scoring-model.md — Externalize that "weight value delivered, relationship strength" thinking with actual examples and formulas
• references/templates.md — Provide the advocate roster spreadsheet template you mention but don't actually show
• references/consent-tracking.md — Detail the compliance workflow separately

This alone gets you to 22/30 on PDA. The point is: developers need templates and scoring weights before they ask follow-ups. Layer it.

Other Things Worth Fixing

1. Add trigger phrases to the description — Currently it says "Use to identify, score, and prioritize customer advocates." Real users search for "score customer advocates" or "build reference pool." Add those exact phrases to your frontmatter description field (+2 points).

2. Make the framework actionable — "Signal Collection – product usage, NPS, expansion" is too abstract. Change it to: "Collect signals: Query CRM for usage metrics → Export NPS scores → Flag expansion accounts → Aggregate community participation data." Checklist format, not prose (+2 points).

3. Add a validation step — The framework ends at Backlog Management with no feedback loop. Add: "Validate: Test scoring on 5 known advocates → Adjust weights if tiers mismatch → Re-run quarterly." This addresses the "no guidance on testing accuracy" hit (-1 point on Utility) (+3 points).

Quick Wins

• Move templates out of prose — Create actual CSV/JSON examples in references/templates/
• Specify scoring weights — Don't just say "weight these things" — show: Value Delivered (40%), Relationship Strength (30%), Storytelling (20%), Risk (-10%)
• Add real trigger phrases — "when asked to 'score customer advocates', 'build reference pool', 'identify champions', 'audit advocacy coverage'"

Hit these and you're looking at ~77-80/100. The structure gaps are your main drag right now.

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.