I took a look at your mastering-gemini-cli skill and wanted to share some thoughts.
Links:
The TL;DR
You're sitting at 97/100, solidly in A territory. This is graded against Anthropic's best practices for skill design. Your strongest area is Ease of Use (24/25) — the metadata, trigger terms, and decision tables are really well done. The weaker spot is Spec Compliance (12/15), which is mostly about missing trigger phrases in your description and a couple of structural things.
What's Working Well
-
Killer metadata and triggers — Your description calls out specific pain points (approval modes, smartEdit, GEMINI.md, MCP integration) that make the skill immediately discoverable. That's the kind of specificity that actually helps people find what they need.
-
Progressive disclosure nailed — Five focused reference files plus three templates and two scripts is chef's kiss organization. You're not dumping everything in one place; users can go deep without drowning in SKILL.md.
-
Practical templates and checklists — The validate-setup.sh script, the Common Mistakes table, and your three ready-to-use templates (GEMINI.md, settings.json, wrapper script) give people actual things they can copy-paste. That's high utility.
-
Scope clarity — You're explicit about what this skill covers (Gemini CLI headless automation, not auth setup or model selection) and what it doesn't. That boundary-setting prevents confusion.
The Big One: Missing TOC in SKILL.md
At 285 lines, SKILL.md should have a table of contents. Right now, new users have to scroll or search to find what they need. It's not a huge deal when the content is well-written (and yours is), but it's low-hanging fruit for discoverability.
Fix: Add a TOC after the Quick Start section with anchor links to: Headless Automation, Approval Modes, Edit Failure Recovery, MCP Integration, Debugging, etc. This is a +1 point gain and makes the skill feel more polished.
Other Things Worth Fixing
-
First-time setup workflow is scattered — You have a Headless Checklist, but new users don't see a clear numbered 1-2-3 setup path. Add a "Getting Started" section with: install CLI → create settings.json → create GEMINI.md → validate with script → run first command. That's +1 point.
-
Description needs more trigger phrases — You've got a solid description, but only 1-2 trigger phrases. Add: "resume gemini execution", "persistent gemini editing", "gemini approval workflow". More triggers = more people finding this when they search. +1 point for spec compliance.
-
Missing run-check-fix loop in debugging — Your Debug section shows flags but doesn't give a clear iterative pattern. Add: "1. Run with --debug, 2. Check for old_string match, 3. Enable smartEdit if mismatched, 4. Retry." That's +1 on utility.
Quick Wins
- Add TOC to SKILL.md (285 lines → needs navigation)
- Flesh out first-time setup (numbered 1-5 workflow at top)
- Add 3-4 more trigger phrases (approval modes, persistent editing, etc.)
- Add run-check-fix debugging loop (explicit iterative pattern)
These four fixes would push you to 100/100 without changing your core content. You've built something solid — just needs a bit of polish.
Checkout your skill here: [SkillzWave.ai](https://skillzwave.ai) | [SpillWave](https://spillwave.com) 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 mastering-gemini-cli skill and wanted to share some thoughts.
Links:
The TL;DR
You're sitting at 97/100, solidly in A territory. This is graded against Anthropic's best practices for skill design. Your strongest area is Ease of Use (24/25) — the metadata, trigger terms, and decision tables are really well done. The weaker spot is Spec Compliance (12/15), which is mostly about missing trigger phrases in your description and a couple of structural things.
What's Working Well
Killer metadata and triggers — Your description calls out specific pain points (approval modes, smartEdit, GEMINI.md, MCP integration) that make the skill immediately discoverable. That's the kind of specificity that actually helps people find what they need.
Progressive disclosure nailed — Five focused reference files plus three templates and two scripts is chef's kiss organization. You're not dumping everything in one place; users can go deep without drowning in SKILL.md.
Practical templates and checklists — The
validate-setup.shscript, the Common Mistakes table, and your three ready-to-use templates (GEMINI.md, settings.json, wrapper script) give people actual things they can copy-paste. That's high utility.Scope clarity — You're explicit about what this skill covers (Gemini CLI headless automation, not auth setup or model selection) and what it doesn't. That boundary-setting prevents confusion.
The Big One: Missing TOC in SKILL.md
At 285 lines, SKILL.md should have a table of contents. Right now, new users have to scroll or search to find what they need. It's not a huge deal when the content is well-written (and yours is), but it's low-hanging fruit for discoverability.
Fix: Add a TOC after the Quick Start section with anchor links to: Headless Automation, Approval Modes, Edit Failure Recovery, MCP Integration, Debugging, etc. This is a +1 point gain and makes the skill feel more polished.
Other Things Worth Fixing
First-time setup workflow is scattered — You have a Headless Checklist, but new users don't see a clear numbered 1-2-3 setup path. Add a "Getting Started" section with: install CLI → create settings.json → create GEMINI.md → validate with script → run first command. That's +1 point.
Description needs more trigger phrases — You've got a solid description, but only 1-2 trigger phrases. Add: "resume gemini execution", "persistent gemini editing", "gemini approval workflow". More triggers = more people finding this when they search. +1 point for spec compliance.
Missing run-check-fix loop in debugging — Your Debug section shows flags but doesn't give a clear iterative pattern. Add: "1. Run with --debug, 2. Check for old_string match, 3. Enable smartEdit if mismatched, 4. Retry." That's +1 on utility.
Quick Wins
These four fixes would push you to 100/100 without changing your core content. You've built something solid — just needs a bit of polish.
Checkout your skill here: [SkillzWave.ai](https://skillzwave.ai) | [SpillWave](https://spillwave.com) 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.