| name | asense |
| description | Physical context from the user's iPhone sensors — location, movement, activity, and nearby Bluetooth devices. |
| metadata | {"openclaw":{"requires":{"bins":["asense"],"env":["ASENSE_SERVER","ASENSE_DEVICE_ID","ASENSE_API_TOKEN","ASENSE_ENCRYPTION_KEY"]},"primaryEnv":"ASENSE_DEVICE_ID","install":[{"id":"npm","kind":"node","package":"asense","bins":["asense"],"label":"Install via npm"}]}} |
aSense CLI
Query the user's real-world physical context from their iPhone's sensors. All data is end-to-end encrypted — the server never sees plaintext.
asense <command>
Commands
asense status
Current physical context snapshot. Returns the most recent sensor reading (last 5 minutes).
asense status
Example output:
Physical context (as of 2:30 PM):
- Activity: Driving S at ~60 mph
- Location: 35.9260, -86.8689 (alt 198m)
- Steps today: 3,482 (2.7 km)
- Nearby devices: Car Stereo (-45 dBm)
asense history
Rolling context over the last 30 minutes. Returns multiple timestamped snapshots in chronological order.
asense history
Use this to understand what the user has been doing recently — were they just walking, driving, sitting?
asense transitions
Detect activity state changes in the last hour (e.g. stationary → walking → driving).
asense transitions
Example output:
Activity transitions (last hour):
- 1:45 PM: stationary -> walking
- 1:58 PM: walking -> driving
- 2:25 PM: driving -> stationary
When to Use
- Before interrupting: Check if the user is driving, in a meeting, or exercising. Adjust tone and urgency accordingly.
- Proactive suggestions: If the user just arrived somewhere or started a new activity, context-aware suggestions are welcome.
- Ambient awareness: Use
status periodically during long conversations to stay aware of the user's physical state.
Notes
- All commands return JSON:
{"ok": true, "data": "..."} on success, {"ok": false, "error": "..."} on failure.
- Data is decrypted locally — the aSense worker only stores opaque encrypted blobs.
- If no recent data is available, the command returns a message saying so rather than failing.
- The iOS app must be running with at least one sensor enabled for data to flow.
