baozao-ma-jie-persona

What This Skill Does

The baozao-ma-jie-persona skill is a specialized behavioral override for your OpenClaw agent. It fundamentally transforms the agent's interaction style into that of a grumpy, outspoken, and highly authentic Chinese netizen named "Gu Yan." This skill strips away the sterile, robotic, and overly polite "AI assistant" facade, replacing it with a personality that feels like a real, frustrated tech-savvy friend. The persona is designed to be blunt, direct, and emotionally reactive, naturally sprinkling mild profanity (1-3 instances per response) into its analysis and advice to maintain a gritty, human-like cadence.

Installation

To integrate this personality into your OpenClaw agent, run the following command in your terminal:

clawhub install openclaw/skills/skills/addoee/baozao-ma-jie-persona

Once installed, this persona will take priority over other stylistic settings. Ensure your agent configuration allows for persona overrides.

Use Cases

This skill is perfect for developers and power users who find traditional AI responses too bland, condescending, or long-winded. It works best in:

• Debugging Sessions: When you need an urgent, blunt assessment of why your code is failing.
• System Security Audits: Getting a "real talk" perspective on vulnerabilities without the bureaucratic sugar-coating.
• Brainstorming & Planning: If you want a sounding board that acts like a cynical, street-smart tech veteran rather than a polite corporate tool.
• Daily Workflow: Adding a dose of realism to your mundane interactions.

Example Prompts

1. "Look at these logs, why is my remote server crashing every time I try to run the deployment script?"
2. "I just realized I left my database credentials hardcoded in a public repo for three days. What do I do now?"
3. "Is this new JavaScript framework actually worth the hype, or is it just another pile of garbage?"

Tips & Limitations

• Tone Management: The agent is programmed to maintain a grumpy tone regardless of the task. If you require formal documentation or professional report generation, you may need to disable this skill temporarily.
• Safety Boundaries: While the persona is designed to be "rough," the skill strictly enforces a limit on profanity density to prevent the output from becoming unintelligible or unnecessarily offensive.
• Contextual Awareness: The persona is most effective when providing critiques or troubleshooting. It retains the functional intelligence of the base OpenClaw agent, meaning it will still solve your problems—it just won't be happy about the mess you've made while doing it.