reward-ops

I was curious what a skill for managing reward operations would look like—turns out you've built something that tackles a genuinely tricky domain, though the execution could use some tightening around the Progressive Disclosure architecture to really sing.

Links:

• GitHub
• Your agentic skill in the SkillzWave marketplace

The TL;DR

You're at 68/100, which is D territory—needs work, but salvageable. This is based on Anthropic's skill best practices. Your strongest area is Spec Compliance (11/15), but Utility (11/20) is dragging you down hard. The gap between what you describe and what you actually deliver is the real problem here.

What's Working Well

• Valid YAML structure – Frontmatter is clean and follows conventions correctly
• Clear operational scope – You've identified a real domain problem (catalog management, fulfillment, vendor coordination)
• Organized framework – The five-point structure (Governance, Fulfillment, Vendor Integration, Compliance, Monitoring) is logical and shows you've thought through the problem space
• grep-friendly formatting – Got a bonus point for structure that's easy to parse

The Big One: Missing Reference Files

Here's what's killing your utility score: you mention templates and runbooks that don't exist. You say "Templates: Reward catalog spreadsheet... Fulfillment flowchart... Incident log + RCA template" but none of these are actually provided. This creates a gap where developers read your skill and think "okay, but where do I start?"

The fix: Create a references/ directory alongside SKILL.md with actual files:

• catalog-template.md – Sample SKU metadata structure
• fulfillment-flowchart.md – Visual or text-based workflow
• incident-rca-template.md – Incident response template

This alone could net you +5 points and make the skill actually useful instead of aspirational.

Other Things Worth Fixing

1. Add trigger phrases to your description – Right now it says "Use to run day-to-day loyalty reward catalog management." Nobody searches for "day-to-day loyalty." Add actual triggers: "reward catalog", "redemption flow", "vendor integration", "points issuance". (+2 points)

2. Turn framework items into actionable workflows – "Catalog Governance – metadata (SKU, cost, eligibility, expiry)" is a topic, not a procedure. Rewrite as: "Launch reward catalog: (a) Define SKU metadata, (b) Configure vendor API, (c) Set eligibility rules, (d) Test in sandbox, (e) Deploy to production." (+6 points)

3. Add validation steps – You mention sandboxes in Tips but no actual run-check-fix patterns. Add a "Validation" section: "After catalog changes: (1) Run test redemption in sandbox, (2) Verify points deduction, (3) Check fulfillment webhook triggered, (4) Confirm vendor order created." (+3 points)

4. Fix the voice – Use imperatives instead of gerunds. Change "Launching or refreshing reward catalogs" to "Launch or refresh reward catalogs." Minor but it makes instructions feel more actionable. (+2 points)

Quick Wins

Focus here for the biggest bang:

• Reference files first – These unlock utility and solve the "but how do I actually use this?" problem
• Trigger phrases – Quick wins for discoverability
• Actionable workflows – Transform your topics into step-by-step procedures
• Validation steps – Show people how to verify their work

You've got the domain knowledge. You just need to package it so developers can actually run with it.

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.