Skip to content

Feedback on your network-engineer skill #5

@RichardHightower

Description

@RichardHightower

I took a look at your network-engineer skill and wanted to share some thoughts.

Links:

The TL;DR

You're at 76/100, solidly in C territory. The grading is based on Anthropic's skill best practices across five pillars. Your Utility scores perfectly at 20/20 — the comprehensive network engineering coverage is genuinely useful. The weak spot is Progressive Disclosure Architecture at 14/30 — that's where you'll get the biggest wins.

What's Working Well

  • Utility is chef's kiss — You've got solid problem-solving power covering cloud architectures, security, and performance. The JSON templates for communication and status reporting are practical and specific.
  • Strong workflow clarity — Your numbered steps and phased approach actually guide people through complex tasks. The checklists for validation are effective.
  • Good trigger discoverability — Terms like "network engineer," "cloud architecture," and "hybrid networks" are baked into the description, making it findable.
  • Dense but organized — For a 253-line file, you've kept consistent terminology and structure throughout. The headers work.

The Big One: Missing Layered Architecture

This is holding you back for a -8 point swing. Right now, you've got one monolithic SKILL.md with everything inlined. What you need is a references directory with detailed guides that the main skill references rather than contains.

Here's the concrete fix:

network-engineer/
├── SKILL.md (trimmed to ~100 lines, just overview + triggers)
├── references/
│   ├── cloud-architectures.md (detailed AWS/GCP/Azure patterns)
│   ├── security-frameworks.md (detailed compliance/hardening)
│   ├── performance-tuning.md (detailed optimization strategies)
│   └── troubleshooting.md (detailed diagnostic procedures)

The main SKILL.md would say "See references/cloud-architectures.md for detailed patterns" instead of listing them all inline. This is token-efficient and scannable. Estimated impact: +8 points, pushing you to 84/100.

Other Things Worth Fixing

  1. Add trigger phrases to your description — Currently reads like a job posting ("Expert network engineer specializing in..."). Change it to: "Performs network engineer operations. Use when asked to 'network engineer', 'run network engineer', or 'network engineer help'." This is a +2 point fix in Spec Compliance.

  2. Ditch the second-person voice — Starting with "You are a senior network engineer" puts readers in second-person perspective. Switch to imperative: "Act as a senior network engineer..." Keeps the voice consistent. Small but noticeable in Writing Style (+2).

  3. Add a table of contents — For 253 lines, a TOC right after the metadata would cut navigation time in half. Just a bulleted list of major sections with anchor links. +1 point for navigation signals.

  4. Tighten the repetitive list structures — You've got similar bullet patterns in multiple sections (Responsibilities, Technical Skills, Specializations). Consolidate where you can and reference the references directory. Improves token economy by ~1 point.

Quick Wins

  • Priority 1: Create references directory structure and refactor content into layered files (+8 points) — this alone gets you to a solid B
  • Priority 2: Add trigger phrases to description (+2 points)
  • Priority 3: Fix voice consistency and add TOC (+3 points)

These three moves bump you from 76 to 89/100.


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.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions