evaluate-skill-descriptions

evaluate-skill-descriptions is a command for GitHub Copilot from dotnet/skills. It costs 18 tokens per session (912 once invoked), scanned A, original, MIT.

A command that rates the descriptions of coding-agent skills from 1 to 5. It checks whether each description explains what the skill does, when to use it, when not to use it, and what makes it distinct.

In plain words
What is it for?
Use it to assess skill descriptions in plugin repositories, especially their triggers, boundaries, outcomes, and distinguishing details.
Why use it?
Poor descriptions can prevent an agent from choosing a relevant skill. The evaluation exposes missing or unclear guidance in skill metadata.

Command for GitHub Copilot

Written for GitHub Copilot: a Copilot chat mode or prompt. Also seen: mentions Claude Code.

Good fit Use it to assess skill descriptions in plugin repositories, especially their triggers, boundaries, outcomes, and distinguishing details.

Compare 6 commands from other repositories ↓
Install with agentmods
npx agentmods add commands/dotnet/skills/evaluate-skill-descriptions
About the project

.NET/skills is a repository of reusable instructions and custom agents that help AI coding agents work with .NET and C#. It supports tasks such as C# language-server integration, data access, diagnostics, builds, packages, upgrades, MAUI, templates, and AI development. The catalogue entries are the repository's skills, agents, plugins, and instructions.

dotnet/skills · 5,383 stars · on GitHub

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Clone the repo
git clone --depth 1 https://github.com/dotnet/skills

Made for: GitHub Copilot.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for evaluate-skill-descriptions

README.md
[![agentmods](https://agentmods.dev/badge/commands/dotnet/skills/evaluate-skill-descriptions.svg)](https://agentmods.dev/commands/dotnet/skills/evaluate-skill-descriptions)
Your own site
<a href="https://agentmods.dev/commands/dotnet/skills/evaluate-skill-descriptions"><img src="https://agentmods.dev/badge/commands/dotnet/skills/evaluate-skill-descriptions.svg" alt="Measured on agentmods" height="20"></a>
Per session 18 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 912 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00018 $0.00912
Opus 5 $0.00009 $0.00456
Sonnet 5 $0.00004 $0.00182
Haiku 4.5 $0.00002 $0.00091

Measured yesterday against content hash 6a469b3813fd, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

evaluate-skill-descriptions scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.github/prompts/evaluate-skill-descriptions.prompt.md · 79 lines

How it starts

The opening of the file, as written. The whole thing — 79 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Evaluate Skill Description Quality

This repo contains several skills for helping AI coding agents like GitHub Copilot and Claude Code with .NET-related coding tasks. Evaluate how good the description is in the YAML front matter for each skill under plugins/**/skills/*/SKILL.md and rate them on a scale of 1–5 where 5 is the best.

What the Description Is For

The description field in each skill's frontmatter is the only information a coding agent sees (alongside the skill name) when deciding whether to load the skill. Agents use progressive loading — they read the full skill content only after they decide the skill is relevant. A poor description means a great skill never gets used.

A good description answers four questions:

  1. What it does — the concrete outcome or capability
  2. When to use it — trigger phrases, scenarios, or user intents that should activate the skill
  3. When NOT to use it — clear boundaries, non-goals, or nearby-but-wrong intents where the skill should be skipped
  4. Key capabilities — enough specifics to differentiate it from other skills

Rating Scale

Rating Meaning
5 Excellent — clearly states what, when to use, when NOT to use, and key capabilities; includes concrete trigger phrases, user-intent signals, and clear non-goals/boundaries
4 Good — covers what and when to use and mentions some boundaries or non-goals, but could be more specific on triggers, edge cases, or capabilities
3 Adequate — describes the skill but is missing clear when-to-use or when-NOT-to-use guidance, lacks strong trigger phrases, or is too generic to reliably match user intent
2 Weak — vague or overly technical; an agent would struggle to know when to activate it or might over-trigger because boundaries and non-goals are unclear
1 Poor — missing, trivially short, or provides almost no actionable information

Examples of Good Descriptions (and Why)

# Good — specific, actionable, with clear boundaries
description: Analyzes Figma design files and generates developer handoff
  documentation. Use when user uploads .fig files, asks for "design specs",
  "component documentation", or "design-to-code handoff". Avoid when user is
  asking about general UX research, copywriting, or non-Figma assets.

# Good — includes trigger phrases and non-goals
description: Manages Linear project workflows including sprint planning, task
  creation, and status tracking. Use when user mentions "sprint", "Linear
  tasks", "project planning", or asks to "create tickets". Avoid when the
  user is managing GitHub issues or other ticketing tools instead of Linear.

# Good — clear value proposition and scope limits
description: End-to-end customer onboarding workflow for PayFlow. Handles
  account creation, payment setup, and subscription management. Use when user
  says "onboard new customer", "set up subscription", or "create PayFlow
  account". Avoid when working with other payment providers or generic CRM
  onboarding flows unrelated to PayFlow.

Read the full file on GitHub · 79 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. yesterday First seen · 79 lines · 18 tokens per session scan A 6a469b3813fd

Subscribe to this mod's changes

evaluate-skill-descriptions is a command published in the GitHub repository dotnet/skills (5,383 stars, last pushed today), licensed MIT. It adds 18 tokens to every session and 912 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-07.