LinkML Schema Reviewer FAQs

Question 1

How does the skill handle different levels of schema violations?

Accepted Answer

It categorizes feedback into three severity levels: ENFORCED (violations causing build failures), ADVISORY (documented conventions), and TECH-DEBT (known issues in existing code).

Question 2

What is the primary purpose of the LinkML Reviewer skill?

Accepted Answer

It provides specialized guidance for reviewing and proposing changes to LinkML schema files (.yaml), specifically tailored for the Alliance of Genome Resources curation model.

Question 3

Can I use this for non-AGR LinkML projects?

Accepted Answer

While it is optimized for the Alliance of Genome Resources (AGR) schema, many of its LinkML best practices and architectural checks are applicable to other biological data modeling projects.

Question 4

Does this skill check for naming conventions?

Accepted Answer

Yes, it advises on naming conventions such as CamelCase for classes, snake_case for slots, and specific suffixes for Data Transfer Objects (DTOs) like _curie or _dto.

Question 5

Does the skill require external data to work?

Accepted Answer

The skill is designed to read local reference files in the repository to understand the current schema inventory, inheritance tree, and import graph before performing a review.

LinkML Schema Reviewer

主要功能

使用场景

LinkML Schema Reviewer

主要功能

使用场景