Endpoints API toolkit with persistent memory (claw-ram/claw-drive) for OpenClaw
npm install openclaw-plugin-endpointsEndpoints API toolkit with persistent memory (claw-ram/claw-drive).
- Scan documents with AI extraction
- Manage categorized endpoints
- Persistent memory (claw-ram index + claw-drive storage)
- OpenClaw transcript support (~95% smaller payloads)
``bash`
openclaw plugins install @openclaw/plugin-endpoints
Or for development:
`bash`
openclaw plugins install -l /path/to/moltbot-skill
Add to your OpenClaw config:
`json`
{
"plugins": {
"entries": {
"endpoints": {
"enabled": true,
"config": {
"apiUrl": "https://endpoints.work",
"apiKey": "ep_your_api_key_here"
}
}
}
}
}
Get your API key from: https://endpoints.work/api-keys
These execute immediately without AI:
| Command | Description |
|---------|-------------|
| /endpoints list | List all endpoints by category |/endpoints inspect /path
| | Get endpoint details |/endpoints stats
| | Show usage statistics |/recall
| | View memory index |/save
| | Instructions for saving to memory |
These are model-invocable tools the AI can use:
| Skill | When It's Used |
|-------|----------------|
| endpoints | Managing endpoints, scanning documents |memory-saver
| | Saving research findings to claw-drive |memory-retriever
| | Fetching prior research from memory |
Two-tier persistent storage:
- claw-ram/index - Quick-reference index with summaries and keywords
- claw-drive/{slug} - Full content storage
Example workflow:
1. User: "Save this research about React patterns"
2. AI invokes memory-saver skillclaw-drive/react-patterns
3. Content saved to
4. Index updated in
5. Later: "What do I know about React?" → AI invokes
`bash``
npm install
npm run typecheck