CtrlK
BlogDocsLog inGet started
Tessl Logo

llm-judge

AI quality judge that scores agent responses 0-10 across helpfulness, accuracy, completeness, and clarity. Use when evaluating multi-agent output or implementing LLM-as-judge quality gates.

71

Quality

88%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

100%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The content is a lean, well-structured instruction skill: concrete output spec, explicit scoring bands, and clear dimensions with no filler. Its only real improvement opportunity is adding a worked calibration example (a sample response with its score and reason) to anchor the scoring bands.

DimensionReasoningScore

Conciseness

The body is lean — no library walkthroughs, no padding, no explanations of concepts Claude doesn't need — and every section (dimensions, output format, guardrails) earns its place; the opening role line is the only mild redundancy and still serves to set context.

3 / 3

Actionability

It gives fully executable guidance: the exact four dimensions to score, a copy-paste JSON output template, the 0-10 integer scale, and concrete scoring bands (0 for empty, 3-5 partial, 6-8 solid, 9-10 exceptional, cap at 8 without justification) — Claude knows exactly what to produce.

3 / 3

Workflow Clarity

This is a simple single-task skill (assess dimensions, apply guardrails, emit JSON) with an unambiguous single action; per the simple-skills note, that qualifies for 3, and no destructive/batch feedback loop is required.

3 / 3

Progressive Disclosure

The skill is under 50 lines with no need for external references and is organized into clearly labeled sections (Skills, Output Format, Guardrails), matching the note that simple skills can score 3 with just well-organized sections.

3 / 3

Total

12

/

12

Passed

Description

77%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is specific and complete, with an explicit 'Use when...' trigger and concrete named scoring dimensions. Its weaker spots are trigger-term naturalness (somewhat technical phrasing) and modest overlap risk with generic evaluation skills.

Suggestions

Broaden natural trigger phrasings a user would actually say, e.g. add 'Use when the user asks to score, grade, rate, or review an agent's response or output quality' to improve trigger-term coverage.

Sharpen distinctiveness by foregrounding the multi-agent coordination context in the trigger clause (e.g. 'evaluating responses from agents in a multi-agent coordination system') to reduce overlap with generic evaluation skills.

DimensionReasoningScore

Specificity

It lists concrete, specific actions — scoring agent responses on a 0-10 scale across four named dimensions (helpfulness, accuracy, completeness, clarity) plus implementing quality gates — matching the 'multiple specific concrete actions' anchor rather than the vague single-domain anchor at 2.

3 / 3

Completeness

It clearly states what the skill does ('scores agent responses 0-10 across...') and gives an explicit 'Use when evaluating multi-agent output or implementing LLM-as-judge quality gates' trigger, satisfying both the what and the when.

3 / 3

Trigger Term Quality

Terms like 'multi-agent output' and 'LLM-as-judge quality gates' are relevant but somewhat technical, and common natural phrasings a user would actually say ('score', 'grade', 'rate', 'review an agent's response') are missing, fitting 'some relevant keywords but missing common variations'.

2 / 3

Distinctiveness Conflict Risk

The multi-agent / LLM-as-judge niche is fairly specific, but 'evaluating output' and 'quality judge' are generic enough that it could still overlap with broader evaluation or quality-scoring skills, fitting 'somewhat specific but could still overlap'.

2 / 3

Total

10

/

12

Passed

Validation

87%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation14 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

metadata_version

'metadata.version' is missing

Warning

metadata_field

'metadata' should map string keys to string values

Warning

Total

14

/

16

Passed

Repository
Atmosphere/atmosphere
Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.