# Alignment Interview

> Establish a shared, decision-ready understanding before implementation by inspecting available evidence and asking one material question at a time. Use when goals, constraints, terminology, or success criteria are unclear; do not use to delay an already approved and sufficiently specified change.

- Skill: `jovanipink/alignment-interview` (Agent Skill, multi-file: 3 files)
- Install (CLI): `npx skillmds@latest add jovanipink/alignment-interview`
- Raw SKILL.md: https://api.skillmd.com/api/skills/jovanipink/alignment-interview/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- License: MIT
- Author: JovaniPink (https://skillmd.com/u/jovanipink)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/jovanipink/alignment-interview

---


# Alignment Interview

Reach decision readiness before implementation starts.

## Workflow

1. Inspect the evidence already available in the request, repository, issue, plan, and named documents before asking the user to repeat facts.
2. Maintain the current understanding as `Goal`, `Users`, `Constraints`, `Known evidence`, `Unknowns`, and `Success criteria`; present only the material change during the interview.
3. Rank unknowns by how much their answers could change scope, architecture, authority, safety, cost, or acceptance.
4. Ask one material question at a time. Explain why it matters and recommend a default when current evidence supports one.
5. Update the shared model after each answer and identify contradictions rather than smoothing them over.
6. Stop when the remaining unknowns do not prevent a responsible next decision. Summarize the agreed model, assumptions, rejected options, unresolved questions, and next authorized action.

## Boundaries

- Do not implement, edit files, publish, or create external records while interviewing unless separately authorized.
- Do not ask the user to repeat facts that can be discovered safely from the authorized workspace.
- Do not treat persistence or question count as quality. Stop when the decision is ready.
- Route broad product discovery to problem framing and concrete implementation design to implementation planning.

## Output

During the interview, return the current material question and changed understanding. At readiness, return `Shared understanding`, `Evidence`, `Decisions`, `Assumptions`, `Open questions`, `Readiness`, and any `Next authorized action`.

## Continuity and evidence

During the interview, present the material question and only the changed understanding needed to answer it. Preserve the full agreed model for the readiness summary; do not repeat it on every turn. Do not reopen decisions already answered unless new evidence creates a material conflict.

Read [the original acceptance example](references/continuity-example.md) when checking this behavior.