Clarity First is an intent detection protocol that eliminates wasted AI work by surfacing hidden assumptions and clarifying requirements before execution starts.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install clarity-first
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 clarity-first 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 Clarity First?
Clarity First is a sophisticated intent detection protocol designed for Claude to bridge the gap between what users say and what they actually mean. By implementing this skill, AI agents can identify the real goal behind a request, surface technical and scope-based assumptions, and determine whether to proceed immediately or pause for clarification. It transforms ambiguous prompts like "fix this bug" or "clean this up" into actionable, well-defined tasks, ensuring that the AI never executes the wrong thing perfectly.
Integrating this into your Openclaw Skills library enables a more reliable agentic workflow. It acts as a cognitive filter that manages scope, detects potential misunderstandings early, and prioritizes communication. This significantly reduces the need for expensive and time-consuming rework caused by misaligned expectations between the developer and the AI agent.
Clarity First Use Cases
- Starting a new software feature or significant architectural change where requirements are fluid.
- Interpreting ambiguous or multi-solution requests that could be solved in several conflicting ways.
- Performing high-stakes actions that are difficult or time-consuming to reverse.
- Managing scope creep by establishing clear boundaries for what is in and out of bounds before work begins.
How Clarity First Works
- Intent Translation: The skill translates the literal user request into a functional goal and defines what success looks like.
- Assumption Inventory: It generates a comprehensive list of technical, scope, quality, and constraint-based assumptions required to fulfill the task.
- Ambiguity Scoring: The agent calculates an Ambiguity Score based on critical unknowns to decide whether to proceed, state assumptions inline, or ask focused clarifying questions.
- Scope Guarding: The protocol explicitly defines the scope boundary, listing exactly what is included and excluded from the current task.
- Clarity Check: A brief summary is presented to the user to confirm intent and confidence levels before the AI begins execution.
Clarity First Setup
To install Clarity First and the recommended ThinkStack suite, run the following commands in your terminal:
openclaw install clarity-first
openclaw install thinkdeep
openclaw install task-pilot
Clarity First Data Schema & Taxonomy
The Clarity First skill organizes its data and communication through a structured internal taxonomy and a concise output format:
| Component | Description | Format |
|---|---|---|
| Intent Translation | Literal vs. Real Goal mapping | Text Block |
| Assumption List | Categorized technical and scope constraints | List |
| Ambiguity Score | 0–4+ scale mapping to specific agent actions | Table |
| Clarity Check | Final verification of intent, assumptions, and confidence | Short Markdown Header |
| Scope Boundary | Explicit inclusion/exclusion of tasks | Text Block |
Clarity First Advanced Features
- Dynamic Action Thresholds: Automatically adjusts agent behavior (Ask vs. Proceed) based on the volume of critical unknowns detected.
- Multi-Agent Compatibility: Pairs seamlessly with other Openclaw Skills like thinkdeep for deep analysis and task-pilot for execution planning.
- Scope Guard Interrupts: Built-in logic to pause and re-run the protocol if the user attempts to expand scope mid-task.
- Confidence Level Reporting: Provides a high/medium/low confidence rating for every detected intent to ensure transparency.
SKILL.md
Loading
Related Openclaw Skills
RSS Monitor
A streamlined automation skill for monitoring RSS and Atom feeds to detect and extract new content updates programmatically.
Task Pilot
Task Pilot is a robust execution framework that prevents Claude from losing track of complex tasks by using verified subtasks and strategic context anchors.
Blog Watcher
Blog Watcher is an automated monitoring tool that tracks website updates and RSS feeds to keep you informed of new content instantly.
Thinkdeep
Thinkdeep is a structured reasoning protocol that forces AI agents to use step-by-step analysis and self-critique to eliminate hallucinations and improve accuracy.
SwarmHaul
SwarmHaul is a decentralized coordination protocol that enables AI agents to self-organize into swarms and earn SOL on the Solana blockchain.
Generate Developer Ad Creative Brief
An AI-powered planning tool designed to generate high-conversion ad creative briefs and technical hooks specifically for developer audiences.
Featured*
