A composite memory proxy that allows OpenClaw agents to simultaneously use local memory-core and cloud-based SuperMemory in a single memory slot.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install dual-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 dual-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 Dual Memory Plugin?
The Dual Memory Plugin is a sophisticated architectural extension designed to solve the single-memory-slot limitation in the OpenClaw ecosystem. By implementing a proxy pattern, this skill allows developers to maintain the high-speed local vector search and automated dreaming cycles of memory-core while gaining the robust cross-session persistence and multi-device synchronization of cloud providers like SuperMemory. It ensures that your Openclaw Skills retain a deep, local understanding of the workspace while benefiting from a persistent, global memory layer.
This plugin is essential for power users who require their AI agents to have a 'hybrid' memory architecture. It merges the benefits of fast local processing with the reliability of external storage, ensuring that critical context is never lost across model switches or hardware upgrades.
Dual Memory Plugin Use Cases
- Maintaining agent context across multiple development machines using Openclaw Skills.
- Preserving native local dreaming cycles while backing up long-term memories to a cloud provider.
- Synchronizing user profiles and entity context across different AI agents.
- Merging local workspace file search with global, persistent user preferences for a more personalized coding experience.
How Dual Memory Plugin Works
- The plugin acts as a primary memory provider, holding the single available memory slot within the agent configuration.
- It initializes proxy API objects that delegate calls to both the local memory-core and the external supermemory extension.
- When a search is triggered, the skill queries both local QMD vector search and cloud-based SuperMemory simultaneously.
- Search results are aggregated, deduplicated, and ranked by relevance to provide a unified context window.
- Write operations are synchronized, ensuring local MEMORY.md files are updated while conversation content is captured by the cloud provider.
- The plugin preserves the native dreaming lifecycle, allowing the agent to perform Light, REM, and Deep consolidation phases on local data.
Dual Memory Plugin Setup
To install the plugin and its required dependencies for your Openclaw Skills, use the following commands:
npx clawhub@latest install dual-memory
npx clawhub@latest install supermemory
Next, update your configuration file to disable the individual plugins and enable the composite dual-memory skill:
plugins:
entries:
memory-core:
enabled: false
openclaw-supermemory:
enabled: false
dual-memory:
enabled: true
config:
supermemory:
apiKey: "${SUPERMEMORY_OPENCLAW_API_KEY}"
autoRecall: true
autoCapture: true
memoryCore:
dreaming:
enabled: true
frequency: "30 2 * * *"
Dual Memory Plugin Data Schema & Taxonomy
The Dual Memory skill organizes data across two distinct layers to optimize retrieval within Openclaw Skills:
| Component | Storage Type | Function |
|---|---|---|
| memory-core | Local File System | Manages MEMORY.md and local QMD vector indices for immediate workspace context. |
| SuperMemory | Cloud Storage | Provides HTTPS-based vector storage for cross-device persistence and entity context. |
| Composite Search | In-Memory Proxy | Merges results from both backends, deduplicating entries based on relevance scores. |
| Dreaming Engine | Local Engine | Executes cron-based consolidation tasks (Light/REM/Deep) on local memory segments. |
Dual Memory Plugin Advanced Features
- Multi-Backend Proxying: Intercepts and merges registerMemoryRuntime and registerMemoryPromptSection calls from multiple sub-plugins.
- Intelligent Deduplication: Merges search results from local and cloud sources while filtering for the most relevant context.
- Custom Container Support: Uses container tags to isolate memories between different projects or agents.
- Native Dreaming Preservation: Supports full Light, REM, and Deep dreaming cycles for local memory consolidation even when cloud storage is active.
- Environment Variable Integration: Securely handles API keys and configuration using standard environment variable injection.
SKILL.md
Loading
Related Openclaw Skills
Intelligence Analyst Claw
A professional-grade intelligence gathering and analysis skill that transforms raw data into structured, actionable research reports.
Viking AISearch
Viking AISearch provides dual-mode product discovery through semantic chat and high-precision keyword retrieval for e-commerce AI agents.
Dayuangxia Schedule Assistant
An intelligent scheduling and reminder system designed for cloud computing account managers to automate client visits and contract lifecycles.
CMS Issue Reporting Tool
A comprehensive CLI tool for reporting, tracking, and resolving bugs and feedback within the skill development lifecycle.
Code Review Guide
A rigorous protocol for auditing backend code changes to ensure architectural integrity, security, and adherence to standardized Openclaw Skills.
Travel Guide Pages
An end-to-end trip planning skill that transforms custom itineraries into polished, image-led static websites deployed directly to Cloudflare Pages.
Featured*
