i18n

I took a look at your i18n skill and wanted to share some thoughts.

Links:

• GitHub
• Your agentic skill in the SkillzWave marketplace

The TL;DR

You're at 76/100, solidly in C territory — that's "adequate with gaps." The skill has strong fundamentals (Spec Compliance scores 14/15), but it's being weighed down by Progressive Disclosure Architecture (18/30). The core issue: everything lives in one 470-line file, so users pay the token cost even for simple queries. Fix that and you're looking at 84-86/100 range.

What's Working Well

• Spec compliance is tight — your YAML frontmatter is valid, naming convention is correct (i18n with proper kebab-case), and your description nails the trigger patterns. You've got 33 different activation patterns covering i18next, locale, RTL/LTR, pluralization — that's comprehensive discoverability.
• Your examples hit the mark — the two detailed input/output samples (pluralization and context-aware translations) show real-world problems and solutions. Users can see exactly what they're getting.
• Terminology is consistent — "i18next," "namespace," "translation" throughout. No confusion about what you're talking about.

The Big One: File Size Architecture

Here's the thing that's holding you back: 470 lines in a single SKILL.md with no progressive disclosure structure. Every query loads the entire file — including those two hefty few-shot examples (174 lines total) — even if someone just needs a quick syntax pattern.

Why it matters: Users asking "how do I set up i18next?" shouldn't load your full pluralization examples. You're burning tokens on content that isn't relevant to that query.

The fix: Create a references/ directory:

• references/patterns.md — move your 8 patterns (namespaces, pluralization, contexts, etc....