A smart interaction protocol that minimizes user friction by asking up to four clarifying questions before executing high-impact or ambiguous tasks.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install guided-conversation
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 guided-conversation 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 Guided Conversation?
The Guided Conversation skill is a sophisticated interaction layer designed to enhance the reliability of autonomous agents within the Openclaw Skills ecosystem. It addresses the common challenge of intent ambiguity by implementing a structured pre-action phase, ensuring that the agent fully understands a user's requirements before performing actions that could be destructive or resource-intensive.
By prioritizing user experience, this skill balances the need for technical precision with the desire for a seamless workflow. It prevents excessive interrogation by strictly limiting the number of follow-up questions, making it an essential tool for developers building high-trust AI systems that require safe execution across file systems, cloud environments, or external APIs.
Guided Conversation Use Cases
- Handling ambiguous user requests that lack critical execution parameters or destination paths.
- Safeguarding high-impact operations such as file deletions, database migrations, or financial transactions.
- Streamlining complex workflows like email generation or automated backups where context is missing.
- Improving the success rate of multi-step agent tasks by validating assumptions early in the lifecycle.
How Guided Conversation Works
- The skill intercepts the user's request during the pre-action hook phase before any tool calls are executed.
- It evaluates the request against current context and memory to identify missing or ambiguous parameters.
- If gaps are found, the agent generates up to 4 targeted clarifying questions to fill those gaps.
- Once sufficient clarity is reached—or the 4-question limit is hit—the agent generates a one-sentence summary of the understood task.
- The agent lists all assumptions made and proceeds to execution, or requests a final confirmation if the task is flagged as high-risk.
Guided Conversation Setup
To integrate this skill into your agent, add it to your manifest and configure the risk parameters. You can initialize the environment using the following command:
openclaw install guided-conversation
Ensure your configuration file includes the following default settings for optimal performance with Openclaw Skills:
# skill.config.json
{
"maxQuestions": 4,
"autoProceedAfterClarity": true,
"requireConfirmationHighRisk": true
}
Guided Conversation Data Schema & Taxonomy
The Guided Conversation skill maintains a structured state to track the interaction lifecycle and minimize repetitive questioning.
| Parameter | Description | Default |
|---|---|---|
maxQuestions |
The maximum number of clarification cycles allowed per task. | 4 |
riskProfile |
Categorization of the task (Low, High) to determine confirmation requirements. | Risk-based |
contextBuffer |
Stores data already provided by the user to avoid redundant questions. | Memory-mapped |
assumptionsList |
A list of inferred values used when the question limit is reached. | Array |
Guided Conversation Advanced Features
- Pre-action hooks that enable the skill to pause any tool execution until safety criteria are met.
- Risk-based conditional logic that overrides automatic progression for sensitive or destructive operations.
- Adaptive context filtering to ensure questions never overlap with information already present in the conversation history.
- Seamless integration with the interaction-pipeline to provide real-time status updates during the clarification process.
SKILL.md
Loading
Related Openclaw Skills
Skill Interview Builder
A structured interview agent that automates the creation of ready-to-use AI skill packages by guiding users through a three-round requirement gathering process.
Interaction Pipeline
A standardized communication protocol that ensures AI agents provide consistent status updates and concise two-line summaries after every action.
FlowForge Workflow Runner
A structured state machine engine for AI agents to execute complex, multi-step workflows without skipping critical steps.
OpenClaw SaaS Builder
A comprehensive orchestration skill for building SaaS projects end-to-end using OpenClaw as the driver and Gemini CLI as a planning copilot.
TradingView Technical Indicators Analysis Assistant
A sophisticated toolset that converts 100+ popular TradingView Pine Script indicators into Python-accessible technical analysis functions.
Amazon SEO
A comprehensive optimization framework for Amazon product listings designed to increase organic rank and conversion rates through strategic keyword placement.
Featured*
