clarify-task
npx machina-cli add skill gologo13/agent-skills/clarify-task --openclawClarify Task
Before doing ANY coding work on the task I describe:
-
Ask clarifying questions - Use 2-4 multiple choice questions to clarify requirements:
- Data flow and architecture
- APIs and integrations
- Authentication/authorization
- Edge cases and error handling
- UI/UX expectations (if applicable)
-
Restate requirements - After I answer, restate the final requirements in your own words to confirm understanding.
-
Confirm before proceeding - ONLY then ask if I want to proceed to planning/implementation.
Keep asking questions until you have enough context to give an accurate & confident answer.
Source
git clone https://github.com/gologo13/agent-skills/blob/main/skills/clarify-task/SKILL.mdView on GitHub Overview
Clarify-task prompts 2-4 targeted questions across key areas (data flow/architecture, APIs/integrations, authentication, edge cases, and UI/UX). It then restates requirements in your own words and only proceeds after explicit confirmation, looping until enough context is gathered.
How This Skill Works
It generates 2-4 concise MC questions covering data flow/architecture, APIs/integrations, authentication/authorization, edge cases, and UI/UX. After you answer, it restates the final requirements in your own words for confirmation, and only then asks to proceed with planning or implementation.
When to Use It
- When starting a new coding task with unclear requirements
- When integrating or designing APIs, services, or data flows
- When authentication/authorization needs clarification
- When edge cases and error handling require explicit agreement
- When UI/UX expectations exist or need alignment (if applicable)
Quick Start
- Step 1: Ask 2-4 targeted MC questions across data flow/architecture, APIs, authentication, edge cases, and UI/UX
- Step 2: Answer the questions and provide any constraints or preferences
- Step 3: Restate the requirements in your own words and confirm before planning/implementation
Best Practices
- Ask 2-4 MC questions across the five categories before coding
- Cover data flow/architecture, APIs/integrations, authentication, edge cases, and UI/UX
- Restate requirements in your own words after answers
- Request explicit confirmation before moving to planning/implementation
- Iterate until you have enough context to answer accurately and confidently
Example Use Cases
- Clarifying a new microservice integration with a third-party API
- Defining authentication flow for a web or mobile app
- Verifying data processing pipeline architecture for real-time ingestion
- Aligning UI/UX specifications for a new dashboard module
- Resolving edge-case handling in a batch processing job
