Cortex for Openclaw

Cortex is a file-based, git-tracked persistent memory system for AI agents that features a behavioral knowledge graph and hybrid search.

waynevaughan
v1.0.0
Feb 24, 2026
0
0
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install cortex-ai

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install cortex-ai 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 Cortex?

Cortex provides a sophisticated infrastructure for AI agents to maintain long-term state and structured recall across multiple sessions. By implementing a dual-layer approach—consisting of the Mind for behavioral memory and the Vault for document storage—this skill allows agents to evolve their understanding of user preferences and complex data environments. As part of the Openclaw Skills ecosystem, Cortex ensures that every memory update is captured as a git commit, providing full transparency and version control for agentic intelligence without requiring constant LLM calls for storage logic.

Cortex Use Cases

  • Storing and reinforcing user preferences and behavioral decisions for more personalized agent interactions.
  • Building a durable document store that agents can query using hybrid search to retrieve contextually relevant data.
  • Implementing automatic memory extraction from live conversations to populate a knowledge graph.
  • Tracking the evolution of agent learning over time through git-integrated history and consolidation routines.

How Cortex Works

  1. The Observer component monitors conversation streams to automatically identify and extract significant entities and patterns.
  2. Data is routed to the Mind for behavioral reinforcement or the Vault for structured document storage.
  3. The system applies specific decay and consolidation algorithms to ensure the most relevant memories are prioritized.
  4. Every write operation is automatically committed to a git repository, creating a permanent, auditable trail of memory changes.
  5. Agents access this stored data using a hybrid search mechanism that blends BM25 keyword matching with semantic and temporal parameters.

Cortex Setup

To integrate this memory layer into your Openclaw Skills workflow, follow these installation steps:

# Clone the Cortex repository
git clone https://github.com/sigmalabs-ai/cortex.git
cd cortex

# Install the necessary dependencies
npm install

# Initialize the memory vault
./cortex init

Cortex Data Schema & Taxonomy

Cortex organizes data into a structured file-based hierarchy optimized for search and retrieval:

Component Description Format
Mind 19 typed entries for behavioral memory, decisions, and lessons JSON/Markdown
Vault Persistent document store with full CRUD capabilities Text/Markdown
Search Hybrid index combining BM25, semantic, and temporal data Vector/File Index
Observer Logic for automatic extraction from conversation logs Process Stream

Cortex Advanced Features

  • Advanced memory decay and reinforcement algorithms to mimic cognitive learning patterns.
  • Seamless git integration for every memory entry, allowing users to roll back or audit agent decisions.
  • Multi-agent support enabling different entities to share or partition their knowledge graphs within the Openclaw Skills framework.
  • Hybrid search functionality that remains performant without the need for a dedicated external database.
  • Zero-LLM dependency for the core storage engine, ensuring data privacy and reducing API latency.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*