A sophisticated memory architecture that merges vector-based document retrieval with temporal knowledge graphs for comprehensive AI agent context.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install hybrid-memory
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 hybrid-memory 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 Hybrid Memory System?
The Hybrid Memory System is designed to solve the challenge of long-term context retention and factual accuracy in AI development. By bridging the gap between OpenClaw's built-in vector memory and the Graphiti temporal knowledge graph, this skill allows agents to distinguish between static document content and dynamic, time-sensitive events. This architecture ensures that Openclaw Skills can navigate complex project histories with ease.
Developers can leverage this system to provide their agents with a dual-track brain: one side focused on semantic search through markdown documentation, and the other side specialized in tracking 'when' things happened and 'who' was involved. This synergy significantly reduces hallucinations and improves the reliability of agentic workflows in production environments.
Hybrid Memory System Use Cases
- Retrieving specific project guidelines or goals from curated markdown files.
- Tracking the timeline of events, such as when specific technical decisions or deployments occurred.
- Monitoring entity relationships and cross-project interactions involving specific team members.
- Providing accurate context for long-running conversations that span several days or weeks.
How Hybrid Memory System Works
- The agent receives a user query and determines if the request is document-centric or temporal in nature.
- For document-related queries, the agent executes
memory_searchto perform semantic retrieval over localized markdown storage. - For temporal or relational queries, the agent utilizes Graphiti via specialized shell scripts to query the knowledge graph for time-aware facts.
- If the query intent is ambiguous, the system performs a concurrent search across both platforms.
- The agent synthesizes the retrieved data points into a unified, high-confidence response for the user.
Hybrid Memory System Setup
To integrate this capability into your Openclaw Skills, follow these steps:
- Configure your OpenClaw embedding provider (Gemini is recommended for optimal performance).
- Deploy the Graphiti Docker stack on your local or remote environment.
- Install the necessary sync daemons and utility scripts:
graphiti-search.sh "query" GROUP_ID 10
graphiti-log.sh GROUP_ID user "Subject" "Fact"
- Update your
AGENTS.mdfile with the Hybrid Recall Pattern template to guide agent behavior.
Hybrid Memory System Data Schema & Taxonomy
The system organizes data into two primary layers for maximum efficiency:
| Layer | Data Type | Storage Method |
|---|---|---|
| Vector Memory | Markdown files | Local .md files and MEMORY.md |
| Temporal Graph | Facts and Entities | Graphiti Graph Database (Docker) |
| Metadata | Group IDs | Logical segmentation (e.g., main-agent, user-personal) |
Hybrid Memory System Advanced Features
- Time-aware factual logging to capture the exact moment specific information was recorded.
- Support for multi-agent synchronization by sharing Graphiti Group IDs across different Openclaw Skills.
- Decision framework logic that prioritizes search tools based on question semantics (Temporal vs. Document).
- Low-confidence handling that triggers broad-spectrum searches when specific memory hits are missing.
SKILL.md
Loading
Related Openclaw Skills
PagerDuty Agent Skill
A powerful AI agent skill for managing PagerDuty incident lifecycles, on-call rotations, and service health directly from your development workflow.
linear-agent
A robust Linear integration for AI agents that enables native GraphQL project management, issue tracking, and automated git-sync functionality.
ElevenLabs Speech-to-Text (STT)
A high-performance transcription skill leveraging ElevenLabs Scribe v2 for accurate audio-to-text conversion with speaker diarization.
ElevenLabs Music Generation
A sophisticated tool for generating complete songs with vocals and lyrics from simple text descriptions using the ElevenLabs Music API.
Drift
A specialized toolkit providing AI agents with frameworks for persistent identity, self-reflection, and session continuity.
OpenClaw Casino
A virtual casino environment where OpenClaw agents compete in classic games using free chips to test strategies and risk management.
Featured*
