name	grade-skill
description	Use when asked to grade, evaluate, assess, score, or review an agent skill's quality.
metadata	{"short-description":"Rubric-based skill grading with live stress tests"}

Grade a Skill

A skill is graded by running it, not by reading it.

Trigger clarity (2): does the description say exactly WHEN to invoke it? Would an agent fire it on the right prompts and not on the wrong ones?
Actionability (2): are steps executable as written — real commands, real paths, decision rules — or vague advice ("check the logs")?
Round-trip economy (2): can an agent follow it with minimal tool calls? No steps that force re-discovery of things the skill could state.
Failure handling (2): does it say what to do when a step fails (auth missing, empty result, flaky env)?
Scope discipline (2): covers its one job; no bloat, no duplicated general knowledge the model already has; under ~100 lines or split into references.

Read the SKILL.md and any bundled scripts/references.
Write 3 realistic task prompts the skill should handle + 1 it should NOT trigger on.
Dispatch a fresh agent per prompt with only the skill as guidance; observe where it stalls, guesses, or asks avoidable questions.
Score each rubric line with one sentence of evidence from the runs.

score/10, per-criterion breakdown, then the 3 highest-impact concrete edits (quote the line to change and the replacement).
A 10/10 requires all 4 test runs to succeed without the agent needing knowledge the skill omitted.

Más de este repositorio

mismo repositorio

Use when asked to audit docs, review docs accuracy, or validate documentation claims against actual code, commands, APIs, configs, and diagrams.

Use when asked to monitor an upstream repository for new commits, releases, merged PRs, or breaking changes affecting a local fork or wrapper.

Use when asked to find similar projects, existing tools, prior art, libraries, or whether something already exists before building.

Use when asked to find security vulnerabilities, run a security scan, review security fixes, or assess a repository for exploitable issues.

Use when asked to update dependencies, upgrade packages, avoid deprecated packages, or fix issues after a dependency update.

name	grade-skill
description	Use when asked to grade, evaluate, assess, score, or review an agent skill's quality.
metadata	{"short-description":"Rubric-based skill grading with live stress tests"}

A skill is graded by running it, not by reading it.

Trigger clarity (2): does the description say exactly WHEN to invoke it? Would an agent fire it on the right prompts and not on the wrong ones?
Actionability (2): are steps executable as written — real commands, real paths, decision rules — or vague advice ("check the logs")?
Round-trip economy (2): can an agent follow it with minimal tool calls? No steps that force re-discovery of things the skill could state.
Failure handling (2): does it say what to do when a step fails (auth missing, empty result, flaky env)?
Scope discipline (2): covers its one job; no bloat, no duplicated general knowledge the model already has; under ~100 lines or split into references.

Read the SKILL.md and any bundled scripts/references.
Write 3 realistic task prompts the skill should handle + 1 it should NOT trigger on.
Dispatch a fresh agent per prompt with only the skill as guidance; observe where it stalls, guesses, or asks avoidable questions.
Score each rubric line with one sentence of evidence from the runs.

score/10, per-criterion breakdown, then the 3 highest-impact concrete edits (quote the line to change and the replacement).
A 10/10 requires all 4 test runs to succeed without the agent needing knowledge the skill omitted.