Prompt file imported from RTSII/GS-Pickleball (
.windsurf/workflows/clarify.md). Fill in{{arguments}}before use. Copyright stays with the author.
The user input to you can be provided directly by the agent or as a command argument - you MUST consider it before proceeding with the prompt (if not empty).
User input:
{{arguments}}
The text the user typed after /clarify in the triggering message is the feature description or clarification request. Assume you always have it available in this conversation even if {{arguments}} appears literally below. Do not ask the user to repeat it unless they provided an empty command.
Given that feature description or clarification request, do this:
-
Load Current Specification: Read the current feature specification from the appropriate spec file to understand the current state and identify areas needing clarification.
-
Structured Clarification Process: Use systematic questioning to clarify underspecified areas, focusing on:
- Functional Requirements: What the feature should do in specific scenarios
- User Experience: How users will interact with the feature
- Technical Constraints: Performance, security, compatibility requirements
- Edge Cases: Error conditions, boundary conditions, and unusual scenarios
- Integration Points: How this feature connects with existing systems
- Success Criteria: How to measure if the feature is working correctly
-
Update Specification: Enhance the specification with clarified requirements, adding or updating sections as needed:
- Add a "Clarifications" section if it doesn't exist
- Update acceptance criteria with clarified requirements
- Add missing user stories or use cases
- Refine technical constraints and non-functional requirements
-
Report Completion: Provide a summary of what was clarified and updated, with the path to the updated spec file and readiness for the
/planphase.
Note: This should be run before /plan to reduce rework downstream. If the user explicitly wants to skip clarification, they should state that clearly.
Clarification Framework:
- Ask about specific user workflows and pain points
- Clarify technical implementation preferences (e.g., "real-time updates or polling?")
- Identify any regulatory or compliance requirements
- Determine performance expectations (e.g., "sub-second response time?")
- Clarify error handling and recovery scenarios