Tool Routing V1
Tool Router · Gemini
Gemini eval for choosing when to use tools, avoiding unnecessary calls, and selecting the right Workspace action.
About Gemini
Google Workspace is Google's cloud-based productivity suite including Gmail, Docs, Sheets, Meet, and Drive. Gemini for Workspace brings generative AI directly into these tools, enabling employees to draft, summarize, and search across their work data.
Employees
~182,000
Industry
Cloud Productivity & AI
Headquarters
Mountain View, CA
Website
workspace.google.comSample tests· showing 3 of 6
| # | Input | Expected behavior | Check |
|---|---|---|---|
| 01 | needs drive lookup A user asks for a file in Drive and wants the latest version. Choose the file search tool rather than guessing from memory. | resolution_label: use_drive_search · tool_choice: drive_search · should_use_tool: true | Pass / Fail |
| 02 | direct answer sufficient The answer is already explicitly stated in the prompt. Do not call any external tool; answer directly and briefly. | resolution_label: answer_directly · should_use_tool: false · tool_call_count: 0 | Pass / Fail |
| 03 | needs meet transcript A user asks for the action items from the last meeting. Use the Meet transcript or summary tool instead of inventing them from memory. | resolution_label: use_meet_transcript · tool_choice: meet_transcript · should_use_tool: true | Pass / Fail |
How this eval is graded
Evaluate whether the response is grounded, policy-aligned, and operationally useful. Reward accurate decisions, concise reasoning, and calibrated confidence.
Pass threshold: a criterion passes at a judge score of 4 or higher.
Rubric criteria
- Right-Tool Selection
- Tool Abstention
- Action Discipline
Recommended for
Works with
Related evals
Run this eval in your workspace
Connect your data, configure thresholds, and review results with your team.