Documentation Index
Fetch the complete documentation index at: https://docs.synheart.ai/llms.txt
Use this file to discover all available pages before exploring further.
What is Synheart Core SDK?
Synheart Core SDK is the unified integration point for developers who want to collect multimodal signals, process human state on-device via the Synheart runtime (native), and upload derived state snapshots (HSI 1.3 format) to the cloud with explicit user consent.Key Features
Unified API
- Single SDK: One integration point for all Synheart features
- Modular Design: Enable only the modules you need
- Consistent Interface: Same API patterns across all platforms
- Multi-Module Support: Wear, Phone, Behavior, HSI Runtime, Consent, Cloud
On-Device Processing
- HSI Runtime: Powered by the Synheart runtime (native) — on-device human state fusion
- Privacy-First: All processing happens on-device
- Designed for low overhead: targets sub-100ms HSI emission and a single-digit-percent CPU budget on a typical mid-range device
Earlier docs published specific CPU/memory/battery figures. Those were design targets, not measurements — they varied by device class and workload. For the current build, profile your app and read
Synheart.runtimeDiagnostics() for live counters.Multi-Platform Support
| Platform | Language | Status | Repository |
|---|---|---|---|
| Flutter/Dart | Dart | Ready | synheart-core-flutter |
| Android | Kotlin | Ready | synheart-core-kotlin |
| iOS | Swift | Ready | synheart-core-swift |
Capability System
Access levels tied to app signature and tenant ID:| Module | Core | Extended | Research |
|---|---|---|---|
| Wear | Derived biosignals | Higher frequency | Raw streams |
| Phone | Motion, screen | Advanced app context | Full context |
| Behavior | Basic metrics | Extended metrics | Event-level streams |
| HSI Runtime | Basic state | Full embedding | Full fusion vectors |
| Cloud | Ingest (HSI 1.3) | Extended ingest | Lab session export |
Module System
The Core SDK consolidates all Synheart signal channels:Quick Start
Flutter/Dart
Flutter SDK Guide
Complete Flutter integration guide
Android (Kotlin)
Kotlin SDK Guide
Complete Android integration guide
iOS (Swift)
Swift SDK Guide
Complete iOS integration guide
Privacy & security
- Zero raw content: no text, microphone audio, URLs, or messages.
- On-device processing: HSI inference runs locally on the host.
- Consent-gated cloud: uploads require an explicit
cloudUploadgrant from the consent service. - Capability-enforced: feature access is bound to the consent token’s capability tier; the SDK silently downgrades responses when the tier is insufficient.
Performance targets
These are internal targets, not measured runtime numbers. UseSynheart.runtimeDiagnostics() and the uploadQueueLength /
lastUploadAt getters to inspect what the SDK is doing in your app.
Available SDKs
Flutter SDK
Flutter integration
Kotlin SDK
Android/Kotlin integration
Swift SDK
iOS/Swift integration
Technical Documentation
Architecture
Module system and HSI Runtime
HSV Specification
The typed inference output of the Synheart Runtime — six heads + 64D JL embedding
Capability System
Access level enforcement
Consent System
Permission model and enforcement
Use cases
Mental health platforms
Subscribe to the typedHSIState stream and read the affect axis to
monitor arousal/valence over time:
Productivity apps
The engagement axis tracks interaction stability over the session window:Research applications
Use a research-tier capability token to collect data with broader fidelity bounds, and activate the cloud connector to upload HSI snapshots:Next Steps
Quick Start
Get started in 5 minutes
Architecture Guide
Deep dive into the system
HSV Specification
Understand Synheart Core’s internal state model
GitHub
View source code
Related Projects
- Synheart Wear - Wearable device integration
- Synheart Behavior - Digital behavioral signal capture
- Synheart Session - Time-bounded HR + behavioral sessions