Install
$
openclaw plugins install clawhub:@getmem/openclaw-getmem@getmem/openclaw-getmem
Persistent memory for OpenClaw agents via getmem.ai.
Every user is remembered across sessions, conversations, and gateway restarts — automatically.
Install
# 1. Install the plugin
openclaw plugins install clawhub:@getmem/openclaw-getmem
# 2. Set your API key
openclaw config set plugins.openclaw-getmem.apiKey gm_live_...
# 3. Restart the gateway
openclaw gateway restart
Get your API key at platform.getmem.ai
How it works
The plugin hooks into two points in the OpenClaw message pipeline:
- Before each LLM call — fetches relevant memory for the user and appends it to the message context automatically
- After each reply — saves the conversation exchange to memory in the background (non-blocking)
No code changes required. Works with all channels (Telegram, Discord, Signal, WhatsApp, etc.)
Configuration
| Key | Required | Default | Description |
|---|---|---|---|
apiKey | ✅ | — | Your getmem.ai API key (gm_live_...) |
baseUrl | — | https://memory.getmem.ai | Custom API endpoint |
enabled | — | true | Disable memory without uninstalling |
Set config values:
openclaw config set plugins.openclaw-getmem.apiKey gm_live_...
openclaw config set plugins.openclaw-getmem.enabled false # to disable
What gets remembered
- User preferences and choices
- Facts about the user's context
- Past decisions and goals
- Relationships and constraints
- Anything mentioned across previous conversations
Memory is scoped per user (by sender ID) and retrieved using semantic search — only relevant context is injected, not everything at once.
Token efficiency
Standard approach: entire conversation history sent every turn (10,000–40,000+ tokens).
With getmem: only relevant memory injected (200–800 tokens). Save up to 95% on context tokens.
Links
- Website: getmem.ai
- Platform (API keys): platform.getmem.ai
- Python SDK: pypi.org/project/getmem-ai
- JavaScript SDK: npmjs.com/package/getmem
License
MIT