coaching-framework

I've been working with coaching frameworks myself, and I'm curious how you're structuring the feedback loops here—the 69 score suggests there's solid foundation but the architecture might be fighting against itself somewhere.

Links:

• GitHub
• Your agentic skill in the SkillzWave marketplace

The TL;DR

You're at 69/100, which puts you in D territory—not broken, but missing some key pieces. This is based on how Anthropic evaluates skill architecture: whether users can actually discover and use it, how well it guides agents through workflows, and whether it delivers real value. Your Writing Style is solid at 8/10, but Progressive Disclosure and Ease of Use are dragging you down at 18/30 and 18/25 respectively.

What's Working Well

• Clear voice consistency: You're using imperative language ("Keep rubrics", "Provide examples") that agents can follow, and you've kept everything focused on practical application without fluff.
• Good naming conventions: coaching-framework is properly formatted, your YAML frontmatter is valid, and the 32-line file is appropriately lean.
• Logical structure: The Framework section with 5 steps and the Tips section flow well—someone can actually understand what you're trying to do.

The Big One: Templates and Examples Are Just Listed, Not Included

This is your biggest miss. You mention three template types (competency card, coaching plan outline, calibration worksheet) but provide zero actual templates. That's like describing what a recipe book should contain without any recipes.

Why it matters: An agent reading this skill knows what to create but not how. The Utility score dropped to 12/20 partly because of this—you're all theory, no concrete working examples.

The fix: Create a `references/t...