architecture-pattern-selector

I spent some time with your architecture-pattern-selector skill and found myself impressed by how you've structured the decision tree—most pattern selection tools either oversimplify or drown you in options, but this one actually guides you through the trade-offs. Your progressive disclosure approach keeps things lean while still delivering real utility, which is exactly why you landed that solid 92.

Links:

• GitHub
• Your agentic skill in the SkillzWave marketplace

The TL;DR

You're at 92/100, A-territory—that's production-ready. This is based on Anthropic's skill best practices. Your strongest pillar is Utility (20/20, perfect score), which makes sense—you're solving a real problem that teams actually face. The weakest area is Spec Compliance (12/15), which is mostly about trigger phrases in your description and some reference organization minutiae.

What's Working Well

• Decision matrix scoring system - You didn't just list patterns, you built an actual evaluation framework with weighted criteria. That's the difference between a reference doc and a tool that prevents bad architecture decisions.
• Layered structure with depth - SKILL.md stays focused while decision-matrix.md and pattern-comparison.md give people the details they need without bloating the overview. That's solid PDA.
• Strong anti-patterns section - Calling out "You're not Netflix" and other reality checks is exactly what engineers need to hear before they cargo-cult architecture decisions.
• Clear 4-step workflow - The numbered selection framework with specific output templates gives people something actionable to follow.

The Big One

Your description is light on trigger phrases. Right now it's pretty generic about "cto-architect" scenarios, but you could pull in more specific triggers that match how people actually describe their problems—like "migrate from monolith", "scale microservices", "modular monolith tradeoffs". This matters because skills get discovered through search and context matching, and you're leaving discovery on the table.

Fix: Add 3-5 more specific trigger phrases to your description that map to real migration and scaling scenarios. Something like: "Use this when you need to select architecture patterns, decide between monolithic vs microservices, evaluate modular monolith viability, plan architecture migration, or analyze service-oriented design trade-offs." That's roughly +1 point and makes the skill way more discoverable.

Other Things Worth Fixing

1. Move references to a dedicated folder - decision-matrix.md and pattern-comparison.md live in the root skill directory, but best practice puts them in references/ to separate overview from details. This is a structural thing that makes the skill feel more organized. (+1 point)

2. Second-person voice in anti-patterns - "You're not Netflix. What's your actual scale?" should be reframed imperatively: "Reality check: Compare actual scale and team size to Netflix's requirements before adopting their architecture." Keeps the tone consistent. (+1 point)

3. TOC for reference files - Since decision-matrix.md and pattern-comparison.md are both substantial docs, add a quick table of contents at the top so people can jump to what they need.

Quick Wins

• Add 3-5 trigger phrases to description (biggest impact—+1 point, way better discoverability)
• Reorganize references into references/ subdirectory (+1 point, cleaner structure)
• Rewrite anti-patterns section to imperative voice (consistency fix, +1 point)
• Add TOCs to reference files (usability improvement)

These changes would push you toward 95+. The skill's already doing real work—this is just making it easier to find and navigate.

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.