A neuroscience-inspired multi-layer memory architecture that replaces flat files with semantic schemas, vector search, and automated consolidation cycles.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install brain-cms
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 brain-cms 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 Brain CMS?
Brain CMS is a sophisticated Continuum Memory System (CMS) designed to revolutionize how AI agents manage long-term information. Rather than relying on a single, bloated memory file that consumes excessive tokens, this system implements a brain-inspired hierarchy. It utilizes semantic schemas, a hippocampal-style router, and a persistent LanceDB vector store to provide sparse, frequency-gated memory loading. As one of the more technical Openclaw Skills, it enables agents to maintain deep context across long-running projects while significantly reducing operational costs.
Brain CMS Use Cases
- Establishing persistent long-term memory for agents handling complex, multi-month projects.
- Reducing token consumption and API costs by replacing flat context injection with sparse schema loading.
- Organizing vast amounts of domain-specific knowledge into retrievable semantic layers.
- Automating memory maintenance through simulated NREM and REM sleep cycles for data consolidation.
How Brain CMS Works
- The agent boots with a lean core context and the most recent daily logs to minimize initial token usage.
- When a specific topic is mentioned, the system consults the INDEX.md router to trigger and load relevant semantic schemas.
- For ambiguous or complex queries, the system performs a semantic similarity search against the LanceDB vector store using local embeddings.
- During the NREM sleep cycle (run on shutdown), the system compresses episodic logs and promotes high-significance events to the ANCHORS.md store.
- During the weekly REM sleep cycle, a local LLM via Ollama synthesizes and consolidates new information into the long-term memory architecture.
Brain CMS Setup
Ensure you have Python 3.10+ and Ollama installed before proceeding with these Openclaw Skills components:
# 1. Run the specialized installer
python3 ~/.openclaw/workspace/skills/brain-cms/install.py
# 2. Index your initial schemas into the vector store
cd ~/.openclaw/workspace/memory_brain
.venv/bin/python3 index_memory.py
# 3. Verify retrieval functionality
.venv/bin/python3 query_memory.py "test topic" --sources-only
Brain CMS Data Schema & Taxonomy
The Brain CMS organizes data across five distinct layers to optimize retrieval and context:
| Layer | Component | Storage Format | Purpose |
|---|---|---|---|
| Working | MEMORY.md | Markdown | Core session context and active tasks |
| Episodic | Daily Logs | Markdown | Daily event tracking and chronological history |
| Semantic | Schemas | Markdown | Permanent domain knowledge and project specs |
| Anchors | ANCHORS.md | Markdown | High-significance ground truth and milestones |
| Vector | LanceDB | Binary/Vector | Semantic search for unstructured data |
Brain CMS Advanced Features
- Automated NREM consolidation that promotes logs tagged with [ANCHOR] to permanent storage.
- Local LLM integration (llama3.2:3b) for free, private memory synthesis and REM sleep cycles.
- Hippocampal routing via INDEX.md to simulate spreading activation for context loading.
- High-performance vector search using LanceDB and nomic-embed-text for fast retrieval.
- Significant token efficiency providing up to 60% reduction in context overhead compared to standard methods.
SKILL.md
Loading
Related Openclaw Skills
Taiwan Surf Point Query
A comprehensive surfing intelligence tool providing real-time wave, tide, and meteorological data for all major surf spots across Taiwan.
Taiwan Real-time Parking Query
An AI-powered skill for querying real-time parking availability and navigation across Taiwan's major cities using TDX data.
B2B Sales Prospecting & Lead Discovery Agent
An automated B2B lead generation tool that identifies high-intent prospects and verified contact data using the Explorium AgentSource API.
Lead & Contact Data Enrichment Agent
A high-performance agent for enriching B2B leads and company profiles with verified contact data, firmographics, and technographics via the Explorium API.
Career Manager
An automated job search pipeline that tracks applications, tailors resumes to job descriptions, and provides deep research for interview preparation.
Lofy Fitness Tracker
A sophisticated fitness accountability layer that logs workouts, meals, and personal records through natural conversation.
Featured*
