A highly actionable and well-sequenced model-selection skill that is let down by verbosity (repeated classifications across matrices, examples, and usage examples) and a broken reference link. Trimming redundant example tables and creating the missing reference.md would materially raise quality.

Suggestions

Create the missing references/reference.md and move the extended/edge-case material there, or remove the broken link from the References section.

Collapse the Examples section (lines 112-161) since it duplicates the decision matrix; keep a few representative cases rather than restating every tier per category.

Consolidate the repeated tier definitions across the four decision matrices and the Philosophy/Anti-Patterns sections to reduce redundancy.

Dimension	Reasoning	Score
Conciseness	The body is comprehensive and domain-specific rather than restating concepts Claude knows, but the large Examples section (lines 112-161) largely reiterates the decision matrix, and the matrices plus examples plus usage examples repeat the same classifications, so it could be tightened considerably.	2 / 3
Actionability	As an instruction/classification skill it provides concrete executable guidance: decision-matrix tables, explicit '+1 tier' escalator rules, a copy-paste output-format template with a worked example, and clear decision rules.	3 / 3
Workflow Clarity	The process is clearly sequenced (parse task -> classify against matrix -> output using format template) with the matrix, escalators, and decision guidance forming an explicit decision pipeline; no destructive/batch validation is needed here.	3 / 3
Progressive Disclosure	It signals a one-level-deep reference for extended detail, but the referenced file references/reference.md does not exist (broken link), and the inline body is a long monolithic block that carries content which could have lived in that reference.	2 / 3
	Total	10 / 12 Passed

Description

90%

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 description with explicit what-and-when structure and natural trigger terms. Its only weakness is that it names a single core action rather than enumerating multiple concrete capabilities.

Dimension	Reasoning	Score
Specificity	The phrase 'Recommend optimal Claude model (haiku/sonnet/opus) for a task' names the domain and the core action (recommend), but offers only one concrete action rather than the multiple specific actions the score-3 anchor calls for.	2 / 3
Completeness	It explicitly states what it does ('Recommend optimal Claude model') and when to use it via a 'Use when...' clause with explicit triggers, satisfying both what and when.	3 / 3
Trigger Term Quality	It includes natural user phrases like "which model", "pick model", and "model for" that a user would actually say when needing this skill.	3 / 3
Distinctiveness Conflict Risk	The model-selection niche with distinct triggers ('which model', 'pick model') is clearly distinguishable and unlikely to fire for unrelated skills.	3 / 3
	Total	11 / 12 Passed

Validation

81%

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 — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
relative_links	Relative link issues: 1 missing	Warning
referenced_paths_exist	Referenced path issues: 1 missing	Warning

	Total	13 / 16 Passed

Reviewed

13 days ago

Table of Contents

Discovery Implementation Validation