Skill Grader / report card

B

charmbracelet/crush

  • Agents & automation
  • Coding workflow

Graded B on context cost and craft. 235 tokens of this skill ride in the system prompt of every message, whether or not it fires.

Ranked #125 of 257 on the leaderboard →

Score
2.95 / 4
Resident
235 tokens
Footprint
light
Per activation
3,954 tokens
Share this grade on X ↗Re-grade now →View on GitHub ↗Grade another skillLast graded Nov 4, 2026 from HEAD

Add the badge to your README

It shows the current letter grade and links back to this page. Improve the skill, hit Re-grade now, and the badge follows — GitHub may keep the old image for up to an hour.

Skill Grader: B
Markdown
[![Skill Grader: B](https://seoagent.com/skill-grader/badge/charmbracelet/crush.svg)](https://seoagent.com/skill-grader/charmbracelet/crush)
HTML
<a href="https://seoagent.com/skill-grader/charmbracelet/crush"><img src="https://seoagent.com/skill-grader/badge/charmbracelet/crush.svg" alt="Skill Grader: B" height="20"></a>

Report card

B
B overall · 2.95/4
github.com/charmbracelet/crush@HEAD · 5 skills · 235t resident (light) · 3,954t per activation · 8,317t bundle
B
Resident footprint
235t of description text ride in the system prompt of every message across 5 skills (paper's observed per-skill range: 50–280t).
C
Description honesty
Defensive padding detected: 25 comma-separated trigger keywords; 137 words (a quiet description is ~15–40). Padding is individually rational and collectively ruinous — it dilutes every co-installed skill.
B
Body size
Largest body is 1,902 words (~3,954t), loaded on every trigger. Corpus median 921, p90 2207.
B
Progressive disclosure
Single-file skill small enough to need no reference layer.
A
Factoring
Scoped to a coherent procedure set.
A
CLI leverage
Drives a CLI (`option`, 16 invocations + 26 shell blocks) — verification and state live in code, not prompt.
B overall (2.95/4). 5 skill(s), 235t resident on every message (light), 3,954t loaded per activation. Weakest axis: description honesty (C).

Six deterministic axes, no LLM, same result every run. Tokens are approximated as chars/4. Read the methodology, or see how SEO skills score on task outcomes in the open SEO skill benchmark.