Issue Clarification

Before diving into code, Otto's PM agent can ask clarifying questions to ensure requirements are well understood.

How It Works

When Otto encounters a ticket without enough detail:

1. Analyzes the ticket — Identifies missing information
2. Posts questions — Adds a comment with specific questions
3. Waits for answers — Pauses until questions are addressed
4. Proceeds with implementation — Once clarified, continues to planning/coding

Example

Original Ticket:

Add export functionality

Otto's Clarifying Questions:

Before I start, I have a few questions:

1. What data should be exportable? (Users, projects, reports?)
2. What format(s) should be supported? (CSV, JSON, Excel?)
3. Should exports be downloadable immediately or sent via email?
4. Are there any size limits or pagination requirements?
5. Should this be available to all users or specific roles?

After Clarification:

Thanks for the details! Here's my understanding:

• Export user list to CSV
• Immediate download (no email)
• Admin role only
• Max 10,000 records per export

I'll proceed with the implementation plan.

Configuring Clarification Behavior

You can control how aggressive Otto is about asking questions in your agent settings:

Always Clarify

Otto asks questions for every ticket, even if requirements seem clear.

Smart Clarification (Default)

Otto only asks when it detects ambiguity or missing information.

Minimal Clarification

Otto proceeds with reasonable assumptions and notes them in the PR.

Best Practices

Enable for New Teams

When Otto is first learning your codebase and conventions, more clarification helps avoid misunderstandings.

Reduce Over Time

As you build a history with Otto and your tickets become more consistent, you can reduce clarification.

Add Context

The more context you provide in your agent configuration, the fewer questions Otto needs to ask.