/oma-deep-interview

/oma:deep-interview

Purpose: Ouroboros-inspired Socratic questioning with mathematical ambiguity scoring. Replaces vague ideas with crystal-clear specifications by iteratively exposing assumptions and gating on ambiguity ≤ 20%.

Usage: /oma:deep-interview <vague idea>

Examples:

• /oma:deep-interview improve the auth system
• /oma:deep-interview --autoresearch make the CI faster
• /oma:deep-interview build a notification system

Pipeline: deep-interview → ralplan (consensus) → ralph/team (execution)

---

How It Works

Ambiguity Dimensions

Dimension	Weight	What It Measures
Goal Clarity	35-40%	Is the primary objective unambiguous?
Constraint Clarity	25-30%	Are boundaries and limitations explicit?
Success Criteria	25-30%	Can acceptance criteria be written as tests?
Context Clarity	15%	(brownfield) Does the change fit the existing system?

Scoring

ambiguity = 1 - weighted_sum_of_dimensions

• ≤ 20% ambiguity: Ready for execution
• 21-50%: Continue questioning
• > 50%: Core concepts still unstable

Challenge Agents

At round 4+: Contrarian Mode — "What if the opposite were true?" At round 6+: Simplifier Mode — "What's the simplest version?" At round 8+: Ontologist Mode — "What IS the core thing here?"

Output

• Spec written to .oma/specs/deep-interview-{slug}.md
• Includes: clarity breakdown, goal, constraints, acceptance criteria, ontology, transcript
• ADR-style decision record with drivers, alternatives, consequences

---

Related

• /oma:interview — lighter-weight Socratic questioning (no ambiguity scoring)
• /oma:ralplan — consensus planning after clarity is achieved
• /oma:autopilot — full pipeline: interview → plan → execute