Prompt file imported from ManisCodeBase/DataHarmony-AI (.github/prompts/speckit.clarify.prompt.md). Copyright stays with the author.

/speckit.clarify

Identify and resolve ambiguities in a feature specification before planning begins.

Instructions

1. Load Feature Spec:

   • Identify target feature from argument (e.g., 004-agentic-hub)
   • Read specs/{feature}/spec.md

2. Analyze for Gaps:

   • Unclear or untestable acceptance criteria
   • Missing edge cases
   • Undefined behavior scenarios
   • Ambiguous terminology
   • Missing non-functional requirements
   • Implicit assumptions not documented

3. Ask Structured Questions (one topic at a time):

   • Provide options where possible to guide answers
   • Focus on ambiguities that will block planning or implementation

4. Update spec.md:

   • Add a "## Clarifications" section with today's date
   • Document each Q&A pair
   • Update affected acceptance criteria or requirements

Output Location

Updates specs/{feature}/spec.md with a Clarifications section

Example Clarification Questions

• "For the live grid polling, should updates appear immediately or with a configurable delay?"
• "When a tool call fails, should the agent retry silently or surface an error event?"
• "Should seed data be deterministic (fixed values) or randomized on each DB reset?"

Templates

constitution.md