A strategic protocol for AI agents to identify ambiguity and seek concise clarification before executing tasks.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install ask-questions-if-underspecified
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install ask-questions-if-underspecified using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Ask Questions If Underspecified?
The Ask Questions If Underspecified skill is a foundational component within the Openclaw Skills ecosystem designed to prevent AI agents from performing "wrong work" due to vague instructions. Instead of making risky assumptions, this skill empowers agents to pause, analyze the request for missing details—such as scope, constraints, and environment—and engage the user with targeted, low-friction questions. By leveraging Openclaw Skills like this, developers can significantly reduce the need for revisions and ensure that every code change aligns perfectly with project requirements.
Ask Questions If Underspecified Use Cases
- When a task lacks specific acceptance criteria or a clear definition of "done."
- During complex refactoring where multiple architectural paths are available.
- When environment details like runtime versions or OS constraints are missing from the prompt.
- Situations where safety, data migration, or reversibility risks need explicit user confirmation through Openclaw Skills protocols.
How Ask Questions If Underspecified Works
- The agent evaluates the request to determine if it is underspecified based on missing objectives, scope, or constraints.
- If ambiguity exists, the agent generates 1-5 scannable, multiple-choice questions to minimize user effort.
- The agent pauses all file modifications and command executions, performing only low-risk discovery like reading configuration files.
- The agent provides a fast-path option, such as a "defaults" response, to streamline the Openclaw Skills interaction.
- Implementation only begins once the agent restates the clarified requirements and receives explicit user confirmation.
Ask Questions If Underspecified Setup
To integrate this logic into your agent workflow, ensure the skill definition is accessible in your configuration path. Many Openclaw Skills can be enabled via the CLI:
# Add the clarification logic to your agent environment
openclaw skill add ask-questions-if-underspecified
Ask Questions If Underspecified Data Schema & Taxonomy
The skill organizes clarification data into several key metadata fields to ensure structured communication within Openclaw Skills:
| Attribute | Purpose |
|---|---|
| Objective | Defines what should change vs. stay the same. |
| Acceptance Criteria | Explicit examples of what constitutes a completed task. |
| Scope | Boundaries for files, components, or user roles. |
| Constraints | Dependencies, performance targets, and style requirements. |
| Safety Logic | Rollout and rollback considerations for risky operations. |
Ask Questions If Underspecified Advanced Features
- Intelligent identification of "Need-to-know" versus "Nice-to-know" information to reduce friction.
- Fast-path response handling allowing users to reply with compact codes (e.g., '1a 2c') or a universal 'defaults' command.
- Low-risk discovery hooks that allow the agent to inspect repo structures before asking questions.
- Integration with wider Openclaw Skills to maintain context across multi-agent sessions.
SKILL.md
Loading
Related Openclaw Skills
Skill Creator
A specialized meta-skill designed to guide developers through the creation, optimization, and packaging of modular capabilities for AI agents.
Technical Translator (CN-EN)
A specialized skill for professional bidirectional technical translation between Chinese and English, ensuring terminology consistency and documentation standards.
Conscious OS — Aligned Voice
A communication framework that transforms AI responses into a calm, precise, and non-reactive coaching voice focused on structural clarity.
Automotive Dealership Marketing
An expert marketing assistant for automotive dealerships designed to optimize inventory merchandising, VDP copy, and OEM-compliant advertising campaigns.
Create Plugin Skill
An automation skill designed to scaffold and implement TypeScript-based plugins, slash commands, and agent tools for the OpenClaw ecosystem.
OpenRouter Usage Tracker
A utility for tracking OpenRouter API spending, credit balances, and per-model cost breakdowns directly from session logs.
Featured*
