The goal is to surface hidden assumptions and force decisions BEFORE implementation begins.
<essential_principles>
</essential_principles>
<quick_start>
- Check if SPEC.md exists and read it
- Identify gaps, ambiguities, and untested assumptions
- Begin interviewing using AskUserQuestionTool
- Cover ALL domains systematically (see question_domains)
- Write completed spec to SPEC.md </quick_start>
If exists: Analyze what's defined vs. what's missing or vague. If not: Start fresh, but ask about existing context (related systems, constraints, prior art).
CRITICAL: Each question must:
- Be specific to THIS project (not generic)
- Surface a decision or assumption
- Have meaningful, distinct options
- Force the user to commit to something
Probe these immediately before moving on.
<question_domains>
</question_domains>
<interview_techniques>
</interview_techniques>
<success_criteria> Interview is complete when:
- All domains have been covered with project-specific questions
- User has made explicit decisions on all tradeoffs
- No "TBD" or "we'll figure it out" remains in critical areas
- Edge cases and failure modes have concrete handling strategies
- The spec could be handed to a developer who would know what to build
- User confirms "this is complete enough to start building" </success_criteria>
<spec_template_location>
See templates/spec-template.md for the output structure.
</spec_template_location>