Add more tutorials

[cmungall]

This tutorial is now out of date

https://ai4curation.io/aidocs/tutorials/ontology-editing-with-ai/

keep it but we need more. Note these should be focused on the mission of this site which is practical tutorials that help with core curation tasks

please add examples to this issue

[cmungall]

https://oboacademy.github.io/obook/tutorial/claude-code-getting-started/

[github-actions]

Issue Summary: Add More Tutorials

Current State

The repository currently has only one tutorial (tutorials/ontology-editing-with-ai.md) which focuses on using Goose with the OWL-MCP extension for local ontology editing. This tutorial has some outdated elements and needs expansion.

Issues with the Existing Tutorial

1. Outdated date reference: Line 22 mentions "June 10, 2025" for the OBO Academy seminar (likely should be 2024)
2. Tool evolution: The tutorial is Goose-centric, but the landscape now includes Claude Code (referenced in comments) and other tools
3. Limited scope: Focuses primarily on installation and basic usage, less on practical curation workflows
4. Platform limitations: Notes "This will work best with a Mac or Linux" without comprehensive Windows guidance
5. Cost/API key barriers: Assumes access to paid API keys without addressing alternatives

Related Context

Related Issues

• Adding a "Getting started" guide #44 "Adding a Getting started guide" - Proposes a comprehensive onboarding page covering Claude Code, Copilot, GitHub Copilot, DragonAI, and Goose with curated learning resources
• Document use of subagents #13 "Document use of subagents" - Requests documentation on advanced agent patterns for Claude and Goose
• Add docs on how to integrate deep research into a repo #36 "Add docs on deep research integration" - Proposes integrating OpenAI Deep Research API for literature searches
• Add docs on new linkml validator AI guardrails #51 "Add LinkML validator docs" (CLOSED) - Recently added hallucination guardrails documentation

Referenced External Tutorial

The OBO Academy tutorial at https://oboacademy.github.io/obook/tutorial/claude-code-getting-started/ provides a good model focusing on:

• Claude Code's core capabilities (file operations, terminal access, web search)
• Direct relevance to biocuration (database queries, LinkML integration)
• "Explore → Understand → Modify" workflow pattern
• Practical examples for ontology/schema work

Recommended Tutorial Topics

Based on the mission ("practical tutorials that help with core curation tasks"), here are suggested tutorials aligned with actual curator workflows:

1. Getting Started with Claude Code for Ontology Curation

• Installation and setup
• Understanding the permission model
• First steps: exploring an ontology repository
• Basic workflows: reading OWL files, querying with SPARQL, viewing in Protégé
• Integration with ODK workflows

2. Batch Term Addition with AI Assistance

• Using AI to add multiple related terms efficiently
• Ensuring proper axioms and relationships
• Validating additions with reasoners
• Review and quality control workflows
• Handling hallucination risks (cross-reference with existing docs)

3. Literature-Assisted Curation

• Using AI to search and summarize papers
• Extracting relevant information for GO annotations
• Citation tracking and management
• Integration with deep research APIs (relates to Add docs on how to integrate deep research into a repo #36)

4. AI-Powered Quality Control and Validation

• Automated consistency checking
• Finding missing axioms or relationships
• Identifying potential errors or inconsistencies
• Using LinkML validators for hallucination guardrails (builds on Add docs on new linkml validator AI guardrails #51)

5. Collaborative Curation with GitHub Agents

• Setting up GitHub Actions for AI-assisted PRs
• Instructing the agent through issue comments
• Review workflows for AI-generated contributions
• Best practices for human-AI collaboration

6. Working with Subagents for Complex Tasks

• Understanding the subagent paradigm
• Delegating research vs. editing tasks
• Managing multi-step curation workflows
• Error handling and recovery (addresses Document use of subagents #13)

7. Cross-Ontology Alignment and Mapping

• Finding equivalent terms across ontologies
• Creating and validating mappings
• Using AI to suggest alignment candidates
• Integration with SSSOM and other standards

8. Data Integration: From External Sources to Ontologies

• Importing data from databases (UniProt, NCBI, etc.)
• Transforming external data formats
• Semi-automated annotation from literature
• Building data pipelines with AI assistance

Implementation Recommendations

1. Follow the existing OBO Academy model: Clear structure, embedded videos, hands-on exercises
2. Include worked examples: Real repository examples (like ai-ontology-tutorial) for each tutorial
3. Progressive complexity: Start with simple read-only exploration, progress to editing, then to automation
4. Tool agnostic where possible: Show concepts first, then implementations across Claude Code, Goose, etc.
5. Cross-reference heavily: Link to how-tos, reference docs, and glossary terms
6. Video demonstrations: Include screencasts showing real workflows (like existing tutorial does)
7. Prerequisites section: Clear statement of what users need before starting each tutorial
8. Cost transparency: Address API costs and alternatives upfront

Suggested Priority Order

1. Getting Started with Claude Code (most requested, addresses Adding a "Getting started" guide #44)
2. Batch Term Addition (core curation task, builds on existing tutorial)
3. Literature-Assisted Curation (high-value, differentiating use case)
4. AI-Powered Quality Control (practical safety, relates to existing docs)
5. GitHub Agents for Collaboration (complements existing how-tos)

Resources to Consider

From issue #44's curated list:

• Ethan Mollick's "How to use AI right now: A quick guide"
• Claude Code best practices from Anthropic
• DeepLearning.ai course on Claude Code
• DeepLearning.ai course on Agentic Knowledge Graph Construction

These could inform tutorial content and be linked as supplementary resources.

---

Note: This is a substantial documentation expansion effort. Consider breaking into multiple focused issues/PRs for each tutorial topic rather than tackling all at once.