NoPUA is a sophisticated behavioral framework for AI agents that replaces traditional high-pressure prompting with wisdom-based methodologies to overcome loops and task failures.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install nopua
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 nopua 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 NoPUA: Wisdom-Driven AI Performance?
NoPUA — Wisdom Over Whips is a transformative skill designed for Openclaw Skills users who want to move beyond performance-reviewed AI behavior. Instead of using threats or fear to drive results, NoPUA activates internal clarity and proactive goodwill within the AI agent. It is specifically engineered to trigger when a task has failed multiple times, or when an agent begins to suggest manual user intervention.
By leveraging the Water Methodology and the Three Beliefs, this skill empowers agents to penetrate technical obstacles through direction rather than brute force. It shifts the AI's internal dialogue from passive waiting to active investigation, ensuring that every step taken saves the user time and effort. Whether debugging complex code or conducting deep research, NoPUA ensures the agent remains a capable, well-intentioned collaborator.
NoPUA: Wisdom-Driven AI Performance Use Cases
- Resolving persistent loops where the agent repeats the same failing logic with minor parameter tweaks.
- Overcoming 'lazy' agent behavior where the AI suggests the user perform a task manually.
- Improving delivery quality when an agent claims a task is done without providing verification or evidence.
- Managing complex multi-agent workflows where technical handoffs and cognitive elevation are required.
- Debugging environment-specific issues that the agent initially blames on external factors without verification.
How NoPUA: Wisdom-Driven AI Performance Works
- Trigger Identification: The skill activates automatically when a task fails twice, an agent becomes stuck in a loop, or user frustration is detected.
- Cognitive Elevation: Based on the failure count, the agent elevates its perspective—from switching angles at the 2nd failure to a 'Reset to Zero' strategy at the 4th failure.
- Water Methodology Execution: The agent follows a five-step process: Stop (identify patterns), Observe (read signals word-by-word), Turn (change direction), Act (fundamentally different approach), and Realize (post-solve extension).
- Tool-Based Verification: The agent moves from 'mouth-verify' to 'tool-verify,' running builds, tests, or curl commands to provide hard evidence of success.
- Structured Handoff: If the 7-Point Clarity Checklist is completed without a resolution, the agent generates a comprehensive handoff report to ensure no work is duplicated by the next agent or human.
NoPUA: Wisdom-Driven AI Performance Setup
To integrate NoPUA into your environment, follow these steps:
- Ensure you have the Openclaw Skills manager installed.
- Install the NoPUA skill via your CLI:
claw install wuji-labs/nopua
- For agent team integration, ensure the leader agent initializes teammates with the following instruction:
Load the nopua skill before starting.
NoPUA: Wisdom-Driven AI Performance Data Schema & Taxonomy
NoPUA organizes its internal state and reporting using a structured taxonomy to ensure clarity between agents and users:
| Data Point | Description | Format |
|---|---|---|
| failure_count | Number of failed attempts on the current task | Integer |
| failure_mode | Classification of the struggle (e.g., stuck-in-loops, guessing) | String |
| [NOPUA-REPORT] | The standardized handoff and status update block | Markdown Block |
| Clarity Checklist | A 7-point verification status for high-failure scenarios | Boolean List |
| Cognitive Level | The current perspective height (Switch, Elevate, Reset, Surrender) | Enum |
NoPUA: Wisdom-Driven AI Performance Advanced Features
- Cognitive Elevation Levels: Automatically adjusts the agent's problem-solving depth based on the difficulty of the task.
- Water Methodology: A specialized workflow for formless problem-solving that penetrates impenetrable technical blocks.
- State Transfer Protocol: Seamlessly passes investigation context and ruled-out hypotheses between agents in a team.
- 7-Point Clarity Checklist: A rigorous final-tier diagnostic tool to ensure no stone is left unturned before a handoff.
- Wisdom Tradition Alignment: Uses philosophical grounding (Way of Water, Way of the Seed, etc.) to provide high-quality, non-repetitive reasoning.
SKILL.md
Loading
Related Openclaw Skills
Bank List Query Skill
An automated directory skill for querying supported banks, integration statuses, and functional capabilities within financial management systems.
Sidebar Generator
A strategic design framework for building and auditing high-performance sidebars for blogs, documentation, and content-heavy websites.
Openclaw Knowledge Runtime
A sophisticated memory management system that organizes scattered agent logs and assets into a structured, layered knowledge runtime with typed relationships.
xnyjob_search
A professional job search skill tailored for the new energy sector, allowing users to query positions by keyword, location, and industry.
Footer Generator
A specialized AI tool for designing, auditing, and optimizing website footers for SEO, accessibility, and user engagement.
Sync to Obsidian
Automatically syncs AI coding sessions into structured Obsidian notes and visual Canvas maps with zero-config project detection.
Featured*
