Architecture
Manifesto's conceptual core is MEL -> Core -> Host.
Do not start here if you are still trying to get the first app running. Use Quick Start and Tutorial first, then come back once the runtime path already feels concrete.
The Current Picture
Manifesto computes deterministic domain state transitions. The system-level mental model is:
MEL source -> Compiler -> DomainSchema
DomainSchema + Snapshot + Intent + Context -> Core.compute()
Core result -> Host effect loop -> terminal Snapshot
SDK -> application-facing actions, reads, observe, inspectMEL declares domain transition rules. Core computes semantic transitions. Host fulfills declared effects and converges snapshots. The SDK exposes that runtime to apps, UIs, backend routes, and agents.
If you need review, audit history, restore, or approval policies, read When You Need Approval or History before adding those extensions.
What Each Layer Owns
| Layer | Responsibility |
|---|---|
| Compiler | MEL to DomainSchema conversion |
| Core | Pure computation of the next semantic state |
| Host | Effect execution and patch application loop |
| SDK | Public application-facing runtime handle |
| Lineage | Optional history, sealing, restore, branch/head history |
| Governance | Optional proposal lifecycle, approval, rejection, decisions |
Read In This Order
- Data Flow
- Determinism
- Failure Model
- When You Need Approval or History
- Lineage Records, only after that decision guide says you need it
When To Read This Section
Read the architecture pages when:
- the tutorial examples make sense, but you want to know why
- you are integrating Manifesto into a larger system
- you need to decide whether to add approval, audit history, or restore
If you are still learning the APIs, go back to Quick Start or Tutorial first.