Issue Summary: Create Gallery of dragon-ai-agent Exemplar Repositories

Overview

This issue requests creating a documentation page showcasing repositories that have implemented the dragon-ai-agent GitHub Actions setup as examples for others to learn from.

Current State & Research Findings

Known Repositories Using dragon-ai-agent

Based on documentation in this repository and related issues, the following ontology repositories are confirmed to use dragon-ai-agent or similar AI agent setups:

Major Ontology Repositories (from Issue #40):

Mondo Disease Ontology - monarch-initiative/mondo
- Has .claude directory and CLAUDE.md
- Referenced as exemplar in docs/how-tos/set-up-github-actions.md:26
Uberon - obophenotype/uberon
- Mentioned in Issue Unify all the different ontology pipelines #40 comments
- Has AI agent setup
Cell Ontology (CL) - obophenotype/cell-ontology
- Has CLAUDE.md file
- Part of the unified ontology pipeline effort
EFO (Experimental Factor Ontology) - Mentioned in Issue Unify all the different ontology pipelines #40
NAMO - Mentioned in Issue Unify all the different ontology pipelines #40

Search Limitations

I attempted to search GitHub's code search API for all repositories with dragon-ai-agent in their .github setup, but don't have permission to use that tool in this context. A manual or alternative search method would be needed to create a comprehensive list.

Current Documentation References

The repository already references several implementation examples:

In docs/how-tos/set-up-github-actions.md:

Links to Mondo's ai-agent.yml workflow as an example
Shows configuration for dragon-ai-agent setup with Goose

In docs/how-tos/instruct-github-agent.md:

Links to Mondo's CLAUDE.md as instruction example
References Uberon PR #3580 for GitHub Copilot setup

Relationship to Issue #40: Pipeline Unification

This gallery request is directly connected to Issue #40: "Unify all the different ontology pipelines". Issue #40 identifies that multiple ontology repositories are using different variants of the semantic coding agent pattern with:

Different instructions
Different agent hosts (Claude Code, Goose, Copilot)
Different deployment strategies
Different billing approaches
Different advanced features

Creating this gallery would serve multiple strategic purposes:

Documentation - Show real-world implementations
Comparison - Enable evaluation of different approaches
Standardization - Support the unification effort in Issue Unify all the different ontology pipelines #40
Knowledge Transfer - Help new repositories adopt best practices
Community Building - Showcase the ai4curation ecosystem

Proposed Gallery Structure

The gallery page could be organized as:

By Agent Platform

Claude Code Implementations
- This repo (ai4curation/aidocs)
- [List others]
Goose Implementations
- Mondo Disease Ontology
- [List others]
GitHub Copilot Implementations
- Uberon (PR #3580)
- [List others]

By Domain/Use Case

Ontology Curation
- Mondo (disease ontology)
- Cell Ontology (cell types)
- Uberon (anatomy)
- EFO (experimental factors)
Knowledge Base Maintenance
Documentation Repositories

Key Information for Each Example

For each repository in the gallery, include:

Repository link and description
Agent platform used (Claude Code, Goose, Copilot)
Key configuration files to examine:
- .github/workflows/ files
- CLAUDE.md or .goosehints
- .config/goose/config.yaml or .claude/settings.json
- .github/ai-controllers.json
Notable features or patterns
Domain-specific adaptations
Link to example PRs created by the agent

Implementation Recommendations

Option 1: Manual Curation (Immediate)

Create a curated list starting with known repositories from this issue and Issue #40, with the expectation to expand over time.

Option 2: Automated Search (Future Enhancement)

Once permissions allow, implement automated GitHub code search to:

search: "dragon-ai-agent path:.github" 
search: "@dragon-ai-agent please" in:file language:yaml

Option 3: Community Contribution

Add a mechanism for repository maintainers to submit their repos to be included in the gallery (e.g., via PR or issue template).

Suggested Page Location

Based on the existing documentation structure:

docs/examples/dragon-ai-gallery.md (new examples section)
OR docs/reference/exemplar-repos.md (in reference section)
OR extend docs/how-tos/set-up-github-actions.md with a comprehensive examples section

Related Issues & Context

Issue Unify all the different ontology pipelines #40 - Unify all the different ontology pipelines (directly related)
Issue Add documentation on use of Claude skills #46 - Add documentation on Claude skills (related to configuration)
Issue Adding a "Getting started" guide #44 - Getting started guide (could link to gallery)
Issue Ensure permissions are set correctly for claude code #39 - Permissions setup (relevant for configuration)

Next Steps

Confirm scope - Should this focus only on dragon-ai-agent or include all AI agent setups?
Gather repository list - Survey community or search programmatically
Create page structure - Decide on organization (by platform, domain, or hybrid)
Write initial content - Start with known examples from Mondo, CL, Uberon
Add to mkdocs nav - Update mkdocs.yml to include the new page
Community feedback - Get input from @dosumis @matentzn @nlharris and others

Value Proposition

This gallery would:

Reduce friction for new adopters by providing proven patterns
Support Issue Unify all the different ontology pipelines #40's unification goals by documenting current state
Enable comparison of different approaches and configurations
Build community by showcasing the growing ai4curation ecosystem
Preserve knowledge about what works in production environments

Create a list of exemplar repos that use ai4curation setup #47

Description

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions