commonmark

I've been digging into markdown parsers lately, and your commonmark implementation caught my eye—solid foundation at 74, though there's definitely room to tighten up the architecture for broader adoption.

Links:

• GitHub
• Your agentic skill in the SkillzWave marketplace

The TL;DR

You're at 74/100, C territory—that's solid fundamentals based on Anthropic's skill best practices. Your strongest area is Spec Compliance (12/15), but you're leaving points on the table with Progressive Disclosure Architecture (21/30) and missing some debugging guidance that would boost Utility from 15/20. The good news? Most of these are fixable without rewriting.

What's Working Well

• Spec compliance is tight - Your YAML frontmatter is clean, name conventions are correct (hyphen-case), and you've got the required fields locked down
• Consistent terminology - You use CommonMark, AST, and node types consistently throughout, which makes the skill easy to follow
• Good trigger coverage for core use cases - "AST structure, block/inline elements, extensions" hits the main search terms developers would use
• Code examples have context - Your two examples show input and output, which beats a lot of skills I've reviewed

The Big One: Missing Validation & Error Handling

This is what's really holding you back on Utility. Right now you're showing the happy path only—what happens when parser.parse() fails? What if someone passes malformed markdown? There's no debugging guidance.

The fix: Add a "Debugging & Edge Cases" section (takes maybe 20 lines):

## Debugging & Edge Cases

- Validate AST with `node.type === 'expected_type'` before accessing properties
- Handle missing children with optional chaining: `node.children?.[0]`
- Empty input returns valid AST with single document node (n...