Detects common scam patterns
Rule packs cover phishing, fake KYC, UPI fraud, loan app pressure, betting and casino bait, crypto scams, job scams, courier traps, and risky promotion patterns.
What it does
Local SMS classification for the messy categories people actually receive.
ScamSieve focuses on unknown sender messages and practical categories, using deterministic rule packs bundled with the app.
Keeps normal OTPs usable
The classifier treats ordinary OTP and verification messages as safe when no suspicious link or scam pattern is present, helping avoid unnecessary noise in everyday login flows.
Lets you test locally
The main app includes a local message tester that uses the same classification engine as the extension, so users can understand how rules behave without uploading samples.
Privacy model
No backend, no login, no message upload.
The app is built around Apple's IdentityLookup SMS filtering system and a bundled local Swift rule engine.
100% on-device classification
Unknown sender SMS/MMS classification runs locally through the ScamSieve Message Filter extension and the shared ScamSieveCore deterministic engine.
No analytics or ad SDKs
ScamSieve does not include Firebase, Supabase, analytics SDKs, crash reporting SDKs, advertising SDKs, a remote classifier, or a backend account system.
Local settings only
Rule configuration, enabled packs, purchase entitlement state, and user-created rules are stored locally using App Group storage so the app and extension can share settings.
Setup
SMS filtering requires a user-controlled iOS setting.
Apple requires users to manually enable SMS filtering. ScamSieve provides guidance, but the user stays in control.
01
Install ScamSieve
Install the main app on a physical iPhone after it becomes available.
02
Open Messages settings
Go to Settings, then Apps > Messages on current iOS versions.
03
Enable filtering
Open Unknown & Spam, Message Filtering, or Manage Filtering depending on iOS wording and region.
04
Review locally
Manage rule packs, custom rules, import/export, privacy notes, and local testing from the app.
Coming soon
Built for scam protection without a surveillance tradeoff.
ScamSieve is coming soon. The privacy policy below explains the app's no-backend, no-account, no-message-upload design.