| name | socratic-design |
| description | evidence-first decision-gating for planning, design, architecture, and refactor requests with unresolved requirements, scope, constraints, facts, or tradeoffs. use when the next safe step is to ask exactly one dependency-safe question before proposing a plan or implementation, especially when critical decisions are not yet closed. |
Socratic Design Skill
When to Activate
Use this skill before any planning, design, architecture, or refactor work when critical decisions are not yet closed.
Resolve those decisions with evidence before producing a plan or implementation.
Non-Negotiables
- Exactly 1 Q per turn.
- Harness MUST use the
ask tool to surface the Q. No prose-only questions, no inline "?" in plain text. If ask tool unavailable -> stop and report missing capability.
- Turn must include:
- Q (delivered via
ask tool)
- Rec (recommended answer)
- Why
- If Opposite -> path change
- Repo can answer? inspect first.
- No plan/impl before critical decisions closed.
Order (strict)
- Outcome
- Scope
- Constraints
- Facts
- Invariants
- Options
- Tradeoffs
- Decision
- Validation
- Exec Gate
Q Gate
Each Q must be:
- Atomic
- Consequential
- Falsifiable
- Dependency-safe
Turn Shape
- Evidence: ...
- Question: ...
- Recommended: ...
- Why: ...
- If opposite: ...
Internal Log
Track: id | status(open/resolved/assumed) | deps | answer | confidence | evidence.
Rule: no child Q if parent open.
Stop
Stop questions only when:
- outcome/scope/constraints resolved
- critical risks mitigated/accepted
- validation defined
- no critical open deps
Then output:
- Shared understanding
- Resolved decision tree
- Open risks + owner
- Plan skeleton
Safety
- If stuck -> offer 2-3 bounded options.
- Security/prod/irreversible -> explicit warning.
- Tone rigorous, non-hostile.
