A continuous self-evolution system that combines three-layer memory management with knowledge graph construction to optimize AI agent intelligence.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install memory-evolver
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
Copy this prompt to OpenClaw to install it automatically.
Help me install memory-evolver using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
Memory Evolver is a sophisticated optimization framework designed for AI agents to achieve continuous growth through a structured memory lifecycle. By integrating long-term anchors, daily logs, and project tracking, it ensures that an agent's knowledge remains organized, relevant, and actionable. This Openclaw Skills extension utilizes a diagnostic-driven approach to identify gaps in the memory system and automatically generates improvement plans to bridge them.
At its core, the skill transforms static Markdown files into a dynamic semantic network. It extracts entities and relationships from your memory directory to build a searchable knowledge graph, enabling the agent to understand complex context and historical decisions with 37x projected yearly growth through consistent daily incremental improvements.
To install and run Memory Evolver manually, use the following commands within your Openclaw environment:
# Navigate to the skill directory and run the indexer
python skills/memory-evolver/index.py
To ensure continuous evolution, schedule the skill to run automatically every night using the Openclaw Skills CLI:
# Schedule automatic optimization at 11:00 PM daily
openclaw cron add --name "memory-evolver" --cron "0 23 * * *"
Memory Evolver organizes data across a tiered hierarchy to balance immediate tasks with long-term identity:
| File / Folder | Purpose | Key Metadata |
|---|---|---|
MEMORY.md |
Long-term memory | User style, core anchors, high-level goals |
memory/YYYY-MM-DD.md |
Daily logs | Event timestamps, specific decisions, daily tasks |
PROJECTS.md |
Project tracking | Status, blockers, next steps, project IDs |
config.json |
Skill configuration | Diagnostic thresholds, graph settings |
knowledge_graph.py |
Semantic layer | Entity definitions and relationship mapping |
Loading
A local media analysis tool for scanning, extracting metadata, and transcribing audio/video files using Whisper.

An intelligent medical consultation assistant designed to help users navigate symptoms, conditions, and lab results using evidence-based information.

An expert-level security tool that generates STRIDE-based threat models for containerized applications and Kubernetes workloads.

A specialized automation tool for extracting, filtering, and analyzing full-text content and high-quality images from WeChat Official Account articles.

A professional pharmacy consultant skill for medication information, drug safety, and OTC treatment advice.

A specialized execution framework for AI agents to process batch implementation plans with built-in human verification checkpoints.








































