I took a look at your build-assistant skill and wanted to share some thoughts.
Links:
The TL;DR
You're at 74/100, solid C-grade territory. This is based on Anthropic's skill best practices rubric. Your strongest area is Writing Style (9/10)—the docs are clear and readable. Weakest is Spec Compliance (10/15)—there are some naming convention issues and missing trigger phrases that are easy wins.
What's Working Well
- Clear documentation structure - Your SKILL.md does a good job layering information: overview → component types → specific workflows. That's PDA done right.
- Practical templates and examples - The skill ships with actual templates for agents, commands, skills, and plugins. That's genuinely useful and shows you thought through what people need.
- Validation scripts - The
validate-*.sh scripts provide real feedback loops. People can run them, check output, fix things. That's excellent utility (18/20).
- Consistent terminology - "component", "manifest", "structure" - you use terms the same way throughout, which makes it easy to follow.
The Big One: Name Convention Issues
Your frontmatter has name: Build-Assistant but it should be lowercase: name: build-assistant. That's a spec compliance issue. Also, when I look at the directory structure, there's a mismatch between the skill name in metadata and where it lives. These seem small, but the spec is strict here because consistent naming makes skills discoverable and parseable by tooling.
Fix: Update line 1 of your SKILL.md frontmatter from Build-Assistant to build-assistant. Check that the directory path matches (should be build-assistant/, not Build-Assistant/).
Impact: +2 points directly.
Other Things Worth Fixing
-
Add a Table of Contents - Your SKILL.md is 122 lines without a TOC. At that length, readers need signposts. Add a quick ## Quick Navigation section right after your description with links to each major section (Installation, Usage, Templates, Validation).
-
Expand trigger phrases - Your description mentions "creating plugin and skill structures" but the "Use when" section only has 1-2 trigger examples. Add more: "bootstrapping new plugins", "creating command stubs", "validating plugin manifests", etc. This helps with discoverability.
-
Clarify documentation references - You reference docs/ files that may not exist in this repo (or are external). Either verify they're there or explicitly note which docs are external to this skill.
Quick Wins
- ✅ Fix name casing to lowercase (
build-assistant)
- ✅ Add TOC for navigation (+3 points, easy)
- ✅ Add 3-4 more trigger phrases to description
- ✅ Verify all referenced docs exist or note if external
These changes alone could push you to 82-85/100 territory (low B-grade). The foundation is solid—it's just polish and spec details holding you back.
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.
I took a look at your build-assistant skill and wanted to share some thoughts.
Links:
The TL;DR
You're at 74/100, solid C-grade territory. This is based on Anthropic's skill best practices rubric. Your strongest area is Writing Style (9/10)—the docs are clear and readable. Weakest is Spec Compliance (10/15)—there are some naming convention issues and missing trigger phrases that are easy wins.
What's Working Well
validate-*.shscripts provide real feedback loops. People can run them, check output, fix things. That's excellent utility (18/20).The Big One: Name Convention Issues
Your frontmatter has
name: Build-Assistantbut it should be lowercase:name: build-assistant. That's a spec compliance issue. Also, when I look at the directory structure, there's a mismatch between the skill name in metadata and where it lives. These seem small, but the spec is strict here because consistent naming makes skills discoverable and parseable by tooling.Fix: Update line 1 of your SKILL.md frontmatter from
Build-Assistanttobuild-assistant. Check that the directory path matches (should bebuild-assistant/, notBuild-Assistant/).Impact: +2 points directly.
Other Things Worth Fixing
Add a Table of Contents - Your SKILL.md is 122 lines without a TOC. At that length, readers need signposts. Add a quick
## Quick Navigationsection right after your description with links to each major section (Installation, Usage, Templates, Validation).Expand trigger phrases - Your description mentions "creating plugin and skill structures" but the "Use when" section only has 1-2 trigger examples. Add more: "bootstrapping new plugins", "creating command stubs", "validating plugin manifests", etc. This helps with discoverability.
Clarify documentation references - You reference
docs/files that may not exist in this repo (or are external). Either verify they're there or explicitly note which docs are external to this skill.Quick Wins
build-assistant)These changes alone could push you to 82-85/100 territory (low B-grade). The foundation is solid—it's just polish and spec details holding you back.
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.