chore: add golden dataset for eval

[Yuan325]

Adding golden datasets that will be used in llm system evaluation.

The golden dataset is separated into multiple types of queries:

• queries that uses a specific tool
• airline related queries (no tool calling - answer is within prompt)
• assistant related question (no tool calling - answer is within prompt)
• out of context questions (no tool calling)
• multitool selections (agent selecting multiple tool before returning a final answer to the user)

[jackwotherspoon]

Just putting this here as a friendly suggestion, can take it or leave it 😄

Long-term open source maintainability can improve drastically by PRs having at a minimum a one or two sentence description (especially for feat PRs but even for all types). Just going over what you are adding and why.

The benefits:

• Helps your reviewer quickly understand the PR
• Helps future maintainers follow-along with PRs
• Allows you to more easily search for old PRs in the future
• Encourages open source external contributions as external users better understand the features and history

I recently have found this extremely beneficial in Cloud SQL repos (example)

Github also recommends this themselves

[kurtisvg]

+1 to Jack's comment. We could do a better job on PR descriptions.