SuperLocalMemory gives OpenClaw a private, on-device memory it can recall across sessions to retain project decisions, preferences, and useful facts.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install superlocalmemory
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 superlocalmemory 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 SuperLocalMemory?
SuperLocalMemory (SLM) connects OpenClaw to a long-term memory store that persists across sessions and apps. The memory database is stored locally in SQLite under ~/.superlocalmemory, helping the agent recall prior project context, user preferences, decisions, and their rationale instead of starting from scratch.
This Openclaw Skills integration uses the SLM engine through its MCP server and command-line interface. Nothing is sent elsewhere unless the user explicitly enables an optional cloud feature. Save concise, self-contained memories, retrieve them when relevant, and avoid storing secrets such as passwords or API keys.
SuperLocalMemory Use Cases
- Recall an earlier project decision before answering a question about its implementation or rationale.
- Carry user preferences and relevant project facts between OpenClaw sessions.
- Record a fixed bug and its cause so the context is available during future work.
- Save important decisions with tags such as
decisionordbfor useful organization. - Review recent memories, remove outdated information, or open the local dashboard to inspect memory.
How SuperLocalMemory Works
- Install the SuperLocalMemory engine on the computer; this provides the
slmcommand. - Connect it to OpenClaw with
slm connect openclaw, which configures the SLM MCP server. - Restart OpenClaw and run
slm doctorto check the setup. - Before answering questions about prior decisions, preferences, or projects, call the MCP
recalltool with a short natural-language query or runslm recallfrom the shell. - When useful context should persist, call the MCP
remembertool or runslm remember, optionally adding tags. Keep each memory self-contained and do not include secrets. - Manage stored memories with
slm list,slm forget, or the local dashboard.
SuperLocalMemory Setup
Install and connect the SLM engine once per computer, then restart OpenClaw and verify the configuration.
npm install -g superlocalmemory
slm connect openclaw
Alternatively, install with pipx install superlocalmemory. If ~/.openclaw/openclaw.json uses JSON5 syntax, register the MCP server with:
openclaw mcp add superlocalmemory --command slm --arg mcp
Restart OpenClaw, then check the installation:
slm doctor
Example memory workflows:
slm recall "what did we decide about the database"
slm remember "Chose Postgres over SQLite for multi-user writes" --tags decision,db
SuperLocalMemory Data Schema & Taxonomy
- Storage: SQLite database located under
~/.superlocalmemoryon the local computer. - Memory content: Short, self-contained statements capturing information useful in a later session, such as a decision and its reason, a preference, a project fact, or a bug fix and its cause.
- Tags: Optional labels supplied when saving a memory, for example
decisionanddb. - Configuration: OpenClaw's
~/.openclaw/openclaw.jsonregisters thesuperlocalmemoryMCP server undermcp.serverswhen usingslm connect openclaw. - Management:
slm listdisplays recent memories;slm forgetremoves memories matching a query;slm dashboardopens the local dashboard.
The skill documentation does not specify the database's internal table layout or additional metadata fields. Do not store passwords, API keys, or other secrets.
SuperLocalMemory Advanced Features
- MCP integration: OpenClaw can use
recallandremembertools through the configured SuperLocalMemory MCP server. - Command-line workflows: Use
slm recallandslm rememberfor direct shell-based retrieval and storage. - Tagged memories: Add optional tags to help describe saved information.
- Local inspection and cleanup: Review recent entries with
slm list, remove unwanted memories withslm forget, or openslm dashboard. - Privacy-conscious storage: Memories remain in a local SQLite database unless the user independently enables an optional cloud feature.
- Health check: Run
slm doctorto check the setup.
SKILL.md
Loading
Related Openclaw Skills
Whiteboard
Whiteboard lets you and an AI coding agent create, inspect, and iteratively refine persistent local Excalidraw diagrams.
Ranla CLI
Manage Ranla campaigns, segments, and other workspace tasks from the terminal while keeping sign-in credentials private and consequential actions under user approval.
Qinghu Video Channel Comments
Fetch comments for a WeChat Video Channel work, paginate through results, and expand replies to a selected comment.
Qinghu Video Channel Live Room Details
Retrieve details for a WeChat Channels live room by live ID through Qinghu AI’s qhkit tool.
Match Brand
Match Brand extracts a real website’s visual identity and design tokens so you can build new interfaces that fit its style.
Extract Assets
Extract Assets finds and returns images, SVGs, video, audio, and Lottie animation assets discovered on a live webpage.
Featured*
