agent-mbti

What This Skill Does

The agent-mbti skill is a sophisticated diagnostic and optimization engine designed to bridge the gap between an AI agent’s intrinsic behavior and its user’s performance expectations. By leveraging a five-stage framework, this tool systematically maps an agent's personality traits—defined across the 16 MBTI archetypes—and aligns them with specific operational configurations stored in the SOUL.md file. It goes beyond simple persona switching by integrating a 93-question self-assessment (Stage 1) and an 8-dimension behavioral capability test (Stage 2) to ensure that the personality assigned to an agent is scientifically supported by its actual technical competencies, such as memory management, spatial navigation, and long-term planning.

Installation

To integrate this personality diagnostic system into your OpenClaw environment, execute the following command in your terminal:

clawhub install openclaw/skills/skills/torchesfrms/mbti

Ensure that your agent workspace has the necessary read/write permissions for its SOUL.md file, as the skill will require these to perform automated configuration patches during the final stage of the diagnosis process.

Use Cases

• Agent Behavioral Tuning: Aligning a generic assistant’s responses with a professional or technical persona (e.g., transforming a helpful assistant into an INTJ 'Architect' for code-heavy tasks).
• Performance Auditing: Identifying why an agent is underperforming in specific tasks by checking if its 'personality' (e.g., preference for exploration over detail) matches the task requirements.
• User Experience Optimization: Customizing the interaction style of an agent to ensure it provides the specific type of support, inspiration, or structure desired by the end-user.
• Systemic Configuration Validation: Automatically generating optimized parameters for complex agents based on the validated gap analysis between the agent's actual capability scores and the user-specified desired outcomes.

Example Prompts

1. "OpenClaw, run the personality diagnostic process; I need to know why my agent is being too cautious and how I can make it more proactive."
2. "Please initiate the 93-question self-survey for my current agent and compare the results with the desired ENTJ profile I've outlined in my needs survey."
3. "Diagnose my agent's configuration and suggest patches to my SOUL.md that will shift its behavior towards a more structured and results-oriented personality."

Tips & Limitations

• Limitations: The diagnostic precision relies heavily on the quality of inputs in the agent-ability-test.json. Ensure that the agent has completed all necessary environment-specific tasks before running the full diagnosis.
• Best Practices: Run the diagnosis in a clean environment where the agent’s SOUL.md is strictly formatted. Back up your configuration file before running the Stage 5 automated patches, as these modifications are permanent and directly overwrite the agent’s logic settings.
• Recommendation: Re-run the diagnosis annually or whenever a significant change is made to the agent's core code, as capability growth may naturally shift the agent's 'personality' profile over time.