Repository navigation
Providers
Find+ polls one or more location providers on a schedule and stores every distinct sighting under a common schema, regardless of source.
Uses the vendored GoogleFindMyTools client. In the Find+ app, Connect signs in
inside a Find+ window (Sign in). From a terminal, findplus auth opens
Chrome for an interactive Google sign-in (undetected-chromedriver, so Google sees a
normal browser sign-in, not an API client). Find+ then stores a long-lived
Android token and the end-to-end-encryption owner key needed to decrypt tag
locations, and polls on your configured interval.
This history consists of locations reported through Google's Find Hub network. Your trackers use nearby participating Android devices to report their location. Location updates can therefore be delayed, sparse, or unavailable, and this application should not be treated as real-time emergency or child-safety GPS tracking.
Optional extra: pip install 'findplus[apple]'. Uses the findmy library
(FindMy.py 0.10). Register an accessory with findplus apple add-accessory,
supplying one of:
- the decrypted pairing record the Find My app keeps for an accessory (a
plist with
privateKey,sharedSecretandpairingDate). FindMy.py parses it and Find+ follows the accessory's rolling keys; - a flat plist with a
Private Keyfield, or a raw base64 private key, for a tag that broadcasts one fixed key. Find My keys are P-224, so the key must be 28 bytes; any other length is refused when you add it.
Genuine Apple AirTags require extracting their pairing keys from an Apple device that has already paired with them; most users cannot do this, so Apple support in practice covers accessories you control the key material for. Locating any accessory also needs an Apple ID sign-in (see Sign in): Apple only answers location queries from a signed-in account.
Apple Find My locations come from nearby Apple devices and can be delayed, sparse or unavailable. Find+ can only query accessories whose keys you hold; genuine AirTags require extracting pairing keys, which most users cannot do.
Each Apple report carries a confidence value (1 to 3) and a one-byte
horizontal accuracy field whose unit neither Apple nor FindMy.py documents.
Neither is a radius Find+ can stand behind, so Find+ does not invent one: an
Apple observation's accuracy is always stored as unknown (null), never a
guessed figure. The dashboard shows "Accuracy unknown" for these fixes instead
of the ±N m reading Google observations carry, and exports leave the
accuracy column empty. Both raw values are kept in the observation's metadata
for anyone who wants them, just never converted into a number.
Each poll asks Apple for the newest report from the last seven days. No report in that window means no new observation; Find+ never fills the gap.
Geofence enter/exit decisions still need some radius to reason about
sparse or missing accuracy. That fallback (geofence_default_accuracy_meters
in Settings) is a documented, conservative constant used only to decide
which side of a place boundary a fix is on. It never becomes the observation's
stored or displayed accuracy.
$ findplus providers
Lists every installed provider and whether it is authenticated and ready to
poll (findplus providers --json for machine-readable output).