Feedback on your using-firebase skill

[RichardHightower]

I took a look at your using-firebase skill and wanted to share some thoughts.

Links:

• GitHub
• Your agentic skill in the SkillzWave marketplace

The TL;DR

You're at 95/100, solid A grade. This is based on Anthropic's skill best practices rubric. Your strongest area is Spec Compliance (14/15) — the metadata and frontmatter are locked in. Weakest area is Utility (17/20) — not because the skill lacks value, but because some promised scripts and templates aren't actually in the package.

What's Working Well

• Progressive Disclosure is chef's kiss. Your SKILL.md is 239 lines with a 3-line Quick Start, then 8 focused references. That's the right balance — dense but navigable. The Task Navigation table with specific triggers like onDocumentCreated, onRequest, onCall actually tells me when to use this.

• Trigger coverage is thorough. You've got 15+ specific keywords in the description — firebase, firestore, cloud functions, callable function, scheduled function, emulator, deploy. Real developers searching for "how do I set up a scheduled Cloud Function" will land here.

• Writing is clean and imperative. No "you should" or "we recommend" — just "Run", "Deploy", "Create". Consistent voice throughout, and the Pre-Deployment Checklist uses checkbox format for actual scanability.

The Big One: Missing Script Implementations

Severity: High — This is what's keeping you from 98+. Your SKILL.md references scripts/init_project.sh, scripts/start_emulators.sh, and scripts/deploy.sh in the Scripts section and Task Navigation table. But they don't exist in the repo.

Why it matters: A developer following your guidance will hit a wall looking for scripts/init_project.sh and won't find it. That breaks trust and makes the skill feel incomplete, even though the reference documentation is solid.

The fix: Either (1) add actual shell scripts to a scripts/ directory with working implementations, or (2) reframe these as example commands you'd run manually and remove the "Scripts" section entirely. If you go route 1, add expected output examples so readers know if things worked. This alone could pull you up +3 points.

Other Things Worth Fixing

1. Assets table references missing files — You mention firebase.json.template and firestore.rules.template but they're not in an assets/ directory. Either add them or cut this section. +2 points if you do.

2. Cross-reference navigation — Your reference files are standalone with TOCs, but they don't link to each other. Add "See also" sections at the end of each reference (e.g., functions-triggers.md should link to functions-patterns.md). Makes it way easier to jump between related topics. +1 point.

3. Scope section is slightly redundant — Lines 39-44 list things "do not use for" but a lot of this is implicit from the skill name. Tighten it to one line: "Out of scope: non-Firebase GCP, other cloud providers, self-hosted backends." +1 point.

Quick Wins

• Add missing scripts/ directory or reframe as CLI commands ✓ +3 points
• Add/include asset templates ✓ +2 points
• Add cross-reference links in references/ ✓ +1 point
• Tighten Scope section ✓ +1 point

Do those four things and you're looking at 99/100. The skill is already strong — these are just gaps between what you documented and what's actually in the package.

---

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.