Eager-load message associations in to_llm to prevent N+1 queries

[matthewbjones]

What this does

ChatMethods#to_llm loads all messages via messages_association.to_a, then iterates calling msg.to_llm on each. Message#to_llm accesses tool_calls, parent_tool_call, and model associations per message — an N+1 query pattern.

This is invisible without strict_loading, but with Rails' strict_loading_by_default = true and :n_plus_one_only mode, every call to conversation.ask raises ActiveRecord::StrictLoadingViolationError.

The same issue exists in cleanup_orphaned_tool_results, which reloads messages and accesses tool_call? / tool_result? on the last message without eager-loading.

Fix: add an eager_load_messages helper that preloads tool_calls, parent_tool_call, and model in a single query. Falls back to plain .to_a when messages_association is not an ActiveRecord relation (e.g. in test stubs).

Type of change

• Bug fix
• New feature
• Breaking change
• Documentation
• Performance improvement

Scope check

• I read the Contributing Guide
• This aligns with RubyLLM's focus on LLM communication
• This isn't application-specific logic that belongs in user code
• This benefits most users, not just my specific use case

Quality check

• I ran overcommit --install and all hooks pass
• I tested my changes thoroughly
• For provider changes: Re-recorded VCR cassettes with bundle exec rake vcr:record[provider_name]
• All tests pass: bundle exec rspec
• I updated documentation if needed
• I didn't modify auto-generated files manually (models.json, aliases.json)

AI-generated code

• I used AI tools to help write this code
• I have reviewed and understand all generated code (required if above is checked)

API changes

• Breaking change
• New public methods/classes
• Changed method signatures
• No API changes

[crmne]

Thanks for your work!