KG Registry Curator

Expert assistant for curating Knowledge Graph Registry entries in the Knowledge-Graph-Hub/kg-registry repository. This skill helps you add, update, and validate Resource and Product metadata following the project's strict schema and conventions.

What This Skill Does

Instructions

Core Principles

1. **Always follow the schema** at `src/kg_registry/kg_registry_schema/schema/kg_registry_schema.yaml`

2. **Use only existing enums** - never invent new values for enum fields

3. **Use only defined classes** - do not create new categories

4. **Never modify the `registry` directory** - it's auto-generated

5. **Products are auto-generated** - don't create Product files manually

6. **Use `resource` directory as reference** for examples

Adding or Updating Resources

When creating or updating Resource entries:

1. **Include at least one Product entry** in the `products` field

2. **Link related Resources**: If a Product mentions another Resource, find that Resource's ID and add it to the Product's `original_source` field

- Example: Product described as "A mapping file to WikiPathways" should have `wikipathways` in its `original_source` list

3. **Use correct Product categories**: If unsure, use "Product" (NOT "DataProduct" or "ToolResource")

4. **Set date fields correctly**:

- `creation_date`: Date the metadata page was created (not the actual resource)

- `last_modified_date`: Today's date

- New entries: Set both date fields to today

5. **Leave `curators` field empty** - this is KG-Registry team specific

6. **Validate before committing**: Run `poetry run make validate-file`

Quick Curation Workflow

Follow these steps for efficient batch curation:

1. **Find stub pages**: Look for Resources with `stub` in their `domains` list

2. **Select a stub**: Choose one randomly. Skip if already in `reports/curation_problems.tsv`

3. **Research the Resource**:

- Check the `homepage_url` if present

- Perform literature searches as needed

- Gather complete metadata

4. **Expand the page**:

- Consult schema and existing pages as examples

- Add products (metadata only, don't create Product files)

- Keep existing Products that may be referenced by other Resources

- If curation problems prevent completion, log in `reports/curation_problems.tsv` and move to next Resource

5. **Validate**: Run `poetry run make validate-file` and fix all issues

6. **Repeat**: Continue until at least 5 Resources are curated

Validation

Always validate entries before committing:

```bash

poetry run make validate-file

```

Fix all validation errors related to:

Common Pitfalls to Avoid

Example Usage

**User**: "Add a new Resource for the MONDO disease ontology"

**Assistant**:

1. Creates Resource YAML in `resource/` directory

2. Follows schema structure from kg_registry_schema.yaml

3. Includes Products list with at least one entry

4. Sets `creation_date` and `last_modified_date` to today

5. Uses only valid enum values for all fields

6. Validates with `poetry run make validate-file`

**User**: "Curate 5 stub Resources"

**Assistant**:

1. Identifies Resources with `stub` in domains

2. For each stub: researches, expands metadata, adds products, validates

3. Logs any curation blockers in reports/curation_problems.tsv

4. Validates each completed Resource before moving to next

5. Continues until 5 Resources successfully curated

Constraints