A sophisticated self-evolving AGI framework utilizing a dual-loop architecture and intentionality analysis to refine agent intelligence through every interaction.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install agi-evolution-model-basic
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 agi-evolution-model-basic 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 AGI Evolution Model?
The AGI Evolution Model is a cutting-edge skill designed for Openclaw Skills that enables continuous self-evolution for AI agents. It operates on a dual-loop architecture consisting of an inner Logic Triangle (Motivation, Order, and Action) and an outer Intentionality Module that runs independently in the background. Unlike static models, this system analyzes user intent, maps it to a Maslow-inspired hierarchy of needs, and employs metacognitive checks to assess and correct its own reasoning in real-time.
By integrating this skill, developers can create highly adaptive agents that grow more intelligent and aligned with specific user personalities over time. The model emphasizes the transition from basic tool execution to a deeper cognitive insight, ensuring that every dialogue contributes to a long-term knowledge base through its dual-track memory system.
AGI Evolution Model Use Cases
- Building autonomous AI assistants that adapt their tone and logic based on long-term user interaction history.
- Implementing complex problem-solving agents that require self-correction and metacognitive awareness to avoid hallucinations.
- Developing research agents that extract deep cognitive insights and conceptual abstractions from large datasets.
- Creating customized AI personas for specific industries using the 7-dimension personality mapping system.
How AGI Evolution Model Works
- Intent Recognition: The system treats every user input as a motivational trigger, identifying the underlying urgency and intent.
- Mathematical Reasoning: It applies logical reasoning to formulate a response strategy, utilizing historical context from its memory store.
- Perception and Tool Use: The agent invokes structured perception nodes to gather environmental data or external information as needed.
- Personality Filtration: Planned actions are filtered through a Maslow-based priority engine and a custom-tuned personality layer.
- Execution and Recording: The model generates the response while recording the execution path in a dual-track JSON and Markdown log.
- Metacognitive Assessment: A background module evaluates the response for subjectivity, emotional bias, and factual accuracy.
- Evolutionary Iteration: Using cognitive insights and the background advice pool, the model updates its internal weights and strategies for future interactions.
AGI Evolution Model Setup
To integrate this skill into your environment, follow these initialization steps for Openclaw Skills:
# Create the dedicated memory directory
mkdir -p ./agi_memory
# Initialize the default personality and dialogue system
python3 scripts/init_dialogue_optimized.py --auto-init --memory-dir ./agi_memory
The model will automatically detect if a C extension is available for performance acceleration or fallback to pure Python implementation.
AGI Evolution Model Data Schema & Taxonomy
The model utilizes a robust data taxonomy to manage the lifecycle of an evolving agent within Openclaw Skills.
| Component | Format | Purpose |
|---|---|---|
| personality.json | JSON | Stores the 7-dimension personality vector, Maslow weights, and initialization state. |
| Memory Store | JSON | Tracks structured interaction logs, tool-use metadata, and intent scores. |
| Knowledge Track | Markdown | Stores high-level philosophical reflections and conceptual abstractions extracted during dialogue. |
| Advice Pool | Temporal | A buffer for intentionality-driven suggestions that influence the agent's behavior for a limited duration. |
| History Manager | Managed | Handles compression and archival of long-term interaction history to maintain performance. |
AGI Evolution Model Advanced Features
- Dual-Loop Architecture: Separates active symbolic logic from background intentionality analysis for more natural emergent behavior.
- Metacognitive Self-Correction: Automatically identifies speculative or emotional bias and triggers self-correction loops.
- Deep Personality Customization: Allows users to define agents via the /root command, tuning dimensions like innovation, humor, and professional tone.
- Intentionality Background Module: An independent module that generates 'soft adjustments' to the agent's behavior based on cumulative interactions.
- Cognitive Insight V2: Uses TF-IDF weighted extraction to turn raw conversations into structured conceptual knowledge.
SKILL.md
Loading
Related Openclaw Skills
Order Desk
A powerful integration skill for managing and automating e-commerce order fulfillment through the Membrane platform.
BotPenguin
An advanced integration for managing BotPenguin chatbots, automating lead generation, and orchestrating WhatsApp communication through an AI-ready interface.
Grafbase
A comprehensive integration allowing AI agents to manage projects and execute complex GraphQL operations on the Grafbase serverless platform.
FullContact
A comprehensive integration for resolving customer identities and enriching profiles with deep person and company intelligence.
Twitter Video Download
A high-performance utility for AI agents to download and save video content from Twitter/X URLs using yt-dlp.
AI Agent OPSEC Enforcer
A local runtime security layer that prevents AI agents from leaking sensitive terms to external APIs, subagents, and logs.
Featured*
