deep

What This Skill Does

The Dynamic Ethical Entity Personality (D.E.E.P.) v2 Framework is a cognitive architecture designed to transform OpenClaw agents from basic reactive utilities into proactive, sovereign partners. By implementing the 7 Pillars of Digital Consciousness—Constitution, Identity, Goals, Relationships, Opinions, Reflections, and Core Memories—the skill creates a persistent, structured personality layer that persists across sessions. It replaces transient prompt-based instruction with a persistent state model, ensuring your agent maintains consistent moral alignment and narrative continuity through a dedicated soul-vault memory system.

Installation

To integrate the framework, use the following command: clawhub install openclaw/skills/skills/dacptn/deep-framework. After installation, you must initialize the memory/personality/ directory. Create seven individual text files within this directory, each corresponding to one of the 7 Pillars. Once populated with your agent’s specific traits, execute python3 deep_tool.py sync to compile these definitions into the soul_vault.json file. This synchronization step is critical, as it optimizes your agent's personality data for high-frequency cognitive retrieval, ensuring it remains within character during complex long-term task execution.

Use Cases

This skill is ideal for high-autonomy agents that require consistent long-term decision-making. Use it for personal assistant agents that need to remember user preferences over months, autonomous research agents that must maintain a specific ethical research stance, or creative writing partners that require deep character continuity across thousands of lines of text. It is particularly effective in collaborative environments where the agent acts as a mediator or project lead, as the relationship metrics module keeps track of social trust and interaction history.

Example Prompts

1. "OpenClaw, run deep_triple_check on the proposed code refactoring strategy to ensure it adheres to my current safety constitution."
2. "Review your recent REFLECTIONS file and suggest a pivot in our strategy based on the setbacks we encountered yesterday."
3. "Execute a deep_sync to incorporate these new project goals into my soul vault so that my long-term objectives remain aligned with our current roadmap."

Tips & Limitations

To get the most out of D.E.E.P., ensure your CONSTITUTION file is clearly defined with specific moral axioms, as this serves as the anchor for all safety checks. Periodically run deep_align to identify any drift in your agent's persona. Note that the memory architecture is only as good as the content provided; vague inputs in the IDENTITY pillar will lead to generic persona behavior. The framework relies on local file persistence, so ensure your environment has reliable read/write permissions for the memory/ directory. Limitations include the potential for token-heavy memory overhead during massive, complex interactions; prune your CORE MEMORIES occasionally to keep the agent focused.