enhanced-memory
An enhanced memory system for OpenClaw agents that replaces the default single-file MEMORY.md with a complete memory architecture: hierarchical directory organization by category, [category:value] tag indexing with multi-tag AND search, automatic lifecycle management (active → archive, never delete), and intelligent cross-category retrieval that auto-routes queries to the right memory module. Gives your agent structured, searchable, long-lived memory out of the box.
Install via CLI (Recommended)
clawhub install openclaw/skills/skills/fatcatmaofei/openclaw-enhanced-memoryEnhanced Memory
A structured memory system that gives your OpenClaw agent organized, searchable, long-lived memory instead of a single monolithic MEMORY.md.
Why?
The default MEMORY.md approach hits a wall fast:
- One file grows endlessly → slow to read, expensive on tokens
- No categorization → food logs mixed with project notes mixed with relationship context
- No retrieval strategy → agent re-reads everything or misses what matters
- No lifecycle → old entries clutter active memory forever
Tagged Memory fixes all of this.
Core Features
1. Hierarchical Directory Organization
Memories are stored in purpose-built directories:
memory/
├── current/ # Active memories (last 6 months)
├── archived/ # Auto-archived older memories (permanent, never deleted)
│ └── YYYY-MM/ # Organized by month
├── RELATION/ # One file per person (relationship context)
├── food/ # Meal and food logs
├── training/ # Exercise and workout records
├── connections.md # Global relationship graph
├── system/ # System config and logs
└── misc/ # Everything else
2. Tag-Based Indexing
Tag any line in any memory file with [category:value] markers:
## 2026-02-20
- Had lunch with Zhang Hao [人物:张浩东] [类型:聚餐] [地点:campus]
- Discussed the new project deadline [项目:openclaw] [类型:会议]
- Yoyo learned a new trick today [宠物:悠悠] [类型:milestone]
Tags support multi-tag AND search — find the exact memory you need:
# Single tag search
python3 scripts/memory_tag_search.py "人物:张浩东"
# Multi-tag AND search (all tags must match)
python3 scripts/memory_tag_search.py "人物:王隆哲" "类型:开票信息"
# List all tags in the system
python3 scripts/memory_tag_search.py --list-tags
# List tags under a specific category
python3 scripts/memory_tag_search.py --list-tags --category 人物
3. Lifecycle Management
Memories age gracefully — never lost, always accessible:
| Age | Location | Status |
|---|---|---|
| 0–6 months | memory/current/ | Active, auto-retrieved |
| 6–12 months | memory/archived/YYYY-MM/ | Archived, searchable on demand |
| 12+ months | memory/archived/ | Permanent archive, manual query |
Run the lifecycle manager manually or via cron:
# Default: archive memories older than 6 months
python3 scripts/memory_lifecycle_manager.py
# Custom threshold (e.g., 90 days)
python3 scripts/memory_lifecycle_manager.py 90
4. Smart Cross-Category Retrieval
The retrieval strategy script auto-classifies queries and searches the right directories:
python3 scripts/memory_retrieval_strategy.py "What did I eat yesterday?"
# → Searches memory/food/ + memory/current/
python3 scripts/memory_retrieval_strategy.py "How is Yoyo doing?"
# → Searches memory/RELATION/悠悠.md + memory/connections.md
python3 scripts/memory_retrieval_strategy.py "Yang Lingxiao"
# → Searches memory/RELATION/ + memory/connections.md
Metadata
Not sure this is the right skill?
Describe what you want to build — we'll match you to the best skill from 16,000+ options.
Find the right skillPaste this into your clawhub.json to enable this plugin.
{
"plugins": {
"official-fatcatmaofei-openclaw-enhanced-memory": {
"enabled": true,
"auto_update": true
}
}
}Related Skills
openclaw-guardian
A security layer plugin for OpenClaw that intercepts dangerous tool calls (exec, write, edit) through two-tier regex blacklist rules and LLM-based intent verification. Critical operations require 3/3 unanimous LLM votes, warning-level operations require 1 LLM confirmation. 99% of normal operations pass instantly with zero overhead. Includes bypass/pipe-attack detection, path canonicalization, SHA-256 hash-chain audit logging, and auto-discovers a cheap model from your existing provider config.
mindcore
Biomimetic emotional mind engine for AI Agents. Provides human-like emotional responses through a 5-layer neural conduction pipeline (L0 Stochastic Noise → L1 Sensor Perception → L2 Subconscious Impulses → L3 Personality Gate → L4 Decision Output) plus 5 psychodynamic patches. Fully decoupled from any LLM — runs locally on CPU with pure Python. Simulates 150 daily impulses across 9 categories with circadian rhythms, mood modulation, and short-term memory influence.