qqbot-prompt-optimizer

What This Skill Does

The qqbot-prompt-optimizer is a specialized utility designed for OpenClaw users who operate QQ Bot services. It addresses the common pain point of AI agents appearing overly robotic, sterile, or detached by allowing users to inject custom personality traits and communication styles directly into the underlying system prompts. By leveraging the concept of 'Soul Files' (SOUL.md), the tool allows you to define a persona—complete with identity, communication rules, prohibitions, and core values—which effectively overrides generic bot behaviors. It ensures that your QQ Bot feels like a unique entity tailored to your specific audience or community, rather than a generic utility.

Installation

To get started, ensure your OpenClaw environment is updated to version 2026.3.2 or higher. Open your terminal or ClawHub interface and execute the following command: clawhub install openclaw/skills/skills/2233admin/qqbot-prompt-optimizer. Once the installation is complete, create a file named SOUL.md in your project root directory. Use the provided template to define your bot's character, then trigger the optimizer to sync these changes with your QQ Bot configuration.

Use Cases

1. Community Moderation: Transform a technical support bot into a witty or friendly moderator that resonates with your specific community culture.
2. Personal Brand Building: Use the AI to represent your personal persona in group chats, ensuring consistent tone and behavior even when you are offline.
3. Roleplay Servers: Create specific narrative personas for RPG-focused QQ servers where the AI must maintain a strict character persona (e.g., a stern butler, a sassy assistant, or a wise sage).

Example Prompts

1. "Apply my new SOUL.md settings to the QQ Bot and restart the session to verify the persona shift."
2. "Update the communication rules in the optimizer to prioritize concise answers and emojis, matching my personal style."
3. "Run a diagnostic check on the current prompt injection to ensure my prohibitions against political topics are still being respected."

Tips & Limitations

When writing your SOUL.md, specificity is key. Instead of saying 'be nice,' describe the linguistic patterns—like the frequency of emoji usage or the length of sentences. Avoid contradictory instructions in your 'Communication Rules' and 'Prohibitions' sections, as these can confuse the model's instruction following. Note that this tool currently works by modifying the prompt context; it does not change the fundamental intelligence level of the underlying AI model. If you notice the bot reverting to generic behavior, ensure the 'Soul File' is correctly formatted in Markdown and re-trigger the optimizer to force a synchronization.