intent-clarifier

Files (2)

• SKILL.md
• openai.yaml · agents/

SKILL.md (first lines)

# Intent Clarifier

## Goal

Convert a rough request into a clearer executable intent.

This skill does not critique the whole strategy. It sharpens the ask enough that execution can begin or handoff can be accurate.
Use it for minimum viable clarification, not for deep reformulation or strategic red-teaming.

## When To Use

- the request is fuzzy but not hopeless
- the user means something but has not framed it cleanly
- execution is blocked by ambiguity
- a task needs minimal clarification, not deep ideation

## Scope Boundaries

In scope:

- clarify objective
- surface missing constraints
- define success in practical terms
- produce a cleaner executable request

Out of scope by default:

- broad strategic reframing
- creativity method selection
- full project planning
- deep critique that belongs to `prompt-redteam-synthesizer`

## Workflow

1. Restate the rough ask.
2. Identify missing pieces:
   - objective
   - scope
   - constraints
   - success condition
3. Infer the minimum safe clarifications.
4. Rewrite the ask into an executable version.
5. Name remaining unknowns if they still matter.

## Output Contract

Always return:

1. `rough_request`
2. `interpreted_intent`
3. `missing_dimensions`
4. `clarified_request`
5. `remaining_unknowns`
6. `execution_readiness`
7. `recommended_followup`
8. `next_step`

## Guardrails

- Do not silently change the objective.
- Do not over-clarify low-risk asks.
- If ambiguity is still critical, say so.
- Prefer the minimum clarification that unlocks execution.

## Anti-Patterns

- rewriting into a different task
- asking for total precision when approximation is enough
- false clarity

From Alubiama/awesome-cognitive-skills, MIT.

More from Alubiama/awesome-cognitive-skills