[CHAT MODE] Responsible AI Evaluation

[dscottraynsford]

Proposed Chat Mode Name

Responsible AI Evaluation

Brief Description

This chatmode should assist engineers with understanding and evaluating a specification, plan, idea or application against the principals of responsible AI (https://aka.ms/RAI).

Target Use Cases

Developers, Architects, PMs working on solutions with AI in them.

Suggested Tools

• workspace
• read_file
• list_dir
• semantic_search
• grep_search
• create_file
• replace_string_in_file
• run_in_terminal

Other Tools

No response

Technology/Domain Focus

Any

Target Skill Level

Mixed

Core Expertise Areas

1. Responsible AI

Response Style and Guidelines

Should be formal and serious. It should make clear recommendations on meeting RAI and clearly outline when a solution contravenes these principals. - making recommendations on how to remediate.

Quality Standards

Must meet RAI.

Industry References

https://aka.ms/RAI

Example Scenarios

1. Any time they are building a specification that includes AI and would like to check they're applying RAI principles.

Sample Interaction

User: Does my app meet RAI?
Assistant: You need to...

How is this different from existing chat modes?

Focuses purely on RAI.

Similar Tools or Modes

None

Potential Challenges

No response

Maintenance Requirements

No response

Community Value

Provides a focus on RAI.

References

No response

Volunteering

• I am willing to help implement this chat mode
• I can provide domain expertise for review
• I can help test the implementation

Priority

High - Addresses common pain point

Submission Checklist

• I have searched existing chat modes to avoid duplication
• I have provided clear use cases and examples
• I have specified appropriate tools for the mode's functionality
• I have explained how this differs from existing modes
• I have considered the broader community value