architecture-design

Use only when creating new registrable ML components that require Factory or Registry patterns.

Quality

53%

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

Fix and improve this skill with Tessl

tessl review fix ./skills/architecture-design/SKILL.md

Quality

Content

50%

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

The body is concrete and well-structured with real code examples, but it is padded with duplicated inline content and ships broken example-file references. Creation workflows lack verification steps and the flagship model example is incomplete.

Suggestions

Remove the inline ~50-line directory tree and point to references/structure.md instead, since that detail already lives there; this also tightens conciseness.

Either create the referenced examples/ files or drop the 'Example Files' section, since all five listed paths currently resolve to nothing.

Complete the model example (define loss/labels/logits instead of 'pass') and add a verification step to the creation workflows, e.g. confirm registration via DATASET_FACTORY / the registry before proceeding.

Dimension	Reasoning	Score
Conciseness	Mostly project-specific and efficient, but padded with generic statements ('modular, extensible architecture with clear separation of concerns'), an 8-bullet 'When Not to Use' list, and a ~50-line directory tree duplicated in references/structure.md.	2 / 3
Actionability	The dataset example is copy-paste ready and the factory/registry snippets are concrete, but the CRITICAL model example uses 'pass' placeholders and references undefined 'loss/labels/logits', and the augmentation section is vague ('Implement transformation function').	2 / 3
Workflow Clarity	Creation workflows are numbered and a code-review checklist exists, but the dataset/model creation steps lack explicit verification checkpoints (e.g., confirm the component registered and is discoverable by the factory).	2 / 3
Progressive Disclosure	The five references/.md files exist and are clearly signaled one level deep, but the full directory structure is duplicated inline (content that belongs in references/structure.md) and every examples/.py path points to a non-existent examples/ directory.	2 / 3
	Total	8 / 12 Passed

Description

57%

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 has a clear, distinctive trigger and niche, but is essentially all 'when' and no 'what' — it never describes the skill's actual capability. Trigger terms lean technical rather than natural user phrasing.

Suggestions

Lead with what the skill does before the trigger, e.g. 'Defines the standard Factory/Registry architecture for ML projects. Use only when creating new registrable ML components (datasets, models, augmentations) that require Factory or Registry patterns.'

Add natural user-facing trigger terms such as 'add a new dataset', 'add a new model', or 'set up a registrable component' alongside 'Factory/Registry patterns'.

Spell out which components qualify ('Dataset, Model, Augmentation, CollateFunction, Metrics') so the trigger is concrete rather than abstract.

Dimension	Reasoning	Score
Specificity	The description names a domain ('registrable ML components') and one action ('creating'), but does not list multiple concrete actions; it stops short of the multi-action anchor.	2 / 3
Completeness	It has an explicit 'Use only when...' trigger (the 'when'), but never states what the skill actually does (e.g., define architecture templates); the 'what' is only implied by the trigger condition.	2 / 3
Trigger Term Quality	'Factory or Registry patterns' are recognizable developer terms, but 'registrable ML components' is technical jargon and common natural phrasing like 'add a new dataset/model' is absent.	2 / 3
Distinctiveness Conflict Risk	The narrow niche ('new registrable ML components that require Factory or Registry patterns') plus 'Use only when' gating makes it unlikely to trigger for unrelated skills.	3 / 3
	Total	9 / 12 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: Galaxy-Dawn/claude-scholar
Path: skills/architecture-design/SKILL.md
Commit: 2f7766f

Reviewed: about 14 hours 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.