advanced-evaluation

This skill should be used when the user asks to "implement LLM-as-judge", "compare model outputs", "create evaluation rubrics", "mitigate evaluation bias", or mentions direct scoring, pairwise comparison, position bias, evaluation pipelines, or automated quality assessment.

Quality

81%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

77%Weight 40%Scale 1-5

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

A highly actionable skill body with excellent prompt templates, worked examples, and validation-equipped workflows. Its weaknesses are token inefficiency from decorative/repeated sections and a monolithic structure with dangling internal references to non-existent files.

Suggestions

Trim the decorative ASCII pipeline diagram and consolidate the numbered Guidelines section, which restates points already covered in Core Concepts and Evaluation Approaches, to reduce token cost.

Split the detailed prompt templates and worked examples into reference files under ./references/ (e.g., prompts.md, examples.md) and link to them from SKILL.md, since the 450-line body is monolithic for a non-simple skill.

Remove or actually create the 'Internal reference' entries (LLM-as-Judge Implementation Patterns, Bias Mitigation Techniques, Metric Selection Guide) — they currently point to files that do not exist in ./references/, creating dangling references.

Dimension	Reasoning	Score
Conciseness	The ~450-line body is mostly substantive but includes padded sections such as the decorative ASCII pipeline diagram and a numbered Guidelines section that restates points already made in Core Concepts and Evaluation Approaches, fitting the 'mostly efficient but could be tightened' anchor rather than the lean 4 or 5.	3 / 5
Actionability	It provides copy-paste-ready prompt templates for direct scoring and pairwise comparison, a concrete numbered position-swap protocol, and three fully worked input/output JSON examples covering the common cases, matching the fully-executable anchor.	5 / 5
Workflow Clarity	The position-bias mitigation protocol is a clearly sequenced multi-step process with an explicit consistency-check validation step and a TIE feedback loop for disagreement, and the anti-patterns section serves as a checklist, matching the explicit-validation-and-feedback anchor.	5 / 5
Progressive Disclosure	The body has good section headers but is a monolithic ~450-line document with everything inlined and no bundle files, and its 'Internal reference' entries (LLM-as-Judge Implementation Patterns, Bias Mitigation Techniques, Metric Selection Guide) point to files that do not exist in ./references/, fitting the 'some structure but content that should be separate is inline' anchor.	3 / 5
	Total	16 / 20 Passed

Description

86%Weight 40%Scale 1-5

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

A strong, trigger-rich description that names a clear niche and uses natural user phrasing throughout. Its only weakness is that capabilities are implied through trigger verbs rather than stated as an explicit 'what this does' clause.

Dimension	Reasoning	Score
Specificity	The description lists several concrete actions embedded in trigger phrasing ("implement LLM-as-judge", "create evaluation rubrics", "mitigate evaluation bias", "compare model outputs"), giving good coverage, but they are framed as triggers rather than explicit capability statements so it falls short of the comprehensive 5 anchor.	4 / 5
Completeness	The "when" is explicit and strong ("This skill should be used when the user asks to... or mentions..."), and the "what" is conveyed through the action verbs in the triggers, but there is no separate explicit capability statement, so it sits between the 4 and 5 anchors rather than clearly answering both.	4 / 5
Trigger Term Quality	It quotes natural phrases users would actually say ("implement LLM-as-judge", "compare model outputs", "create evaluation rubrics") plus synonyms and variations (direct scoring, pairwise comparison, position bias), matching the comprehensive-coverage anchor.	5 / 5
Distinctiveness Conflict Risk	It carves a clear niche (LLM-as-judge evaluation) with distinct triggers like pairwise comparison and position bias, and only minimal overlap with a foundational evaluation skill, matching the clear-niche anchor.	5 / 5
	Total	18 / 20 Passed

Validation

93%

Warnings & errors only

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

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: sickn33/antigravity-awesome-skills
Path: skills/advanced-evaluation/SKILL.md
Commit: 782f684

Reviewed: ~1 hour ago

Table of Contents

Discovery Implementation Validation

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.