Intropy Framework
Read this when a tutorial reaches pipeline steps, context, result variants, or finalizers.
Intropy Framework is the .NET library most Components are built with. Its core idea is the pipeline: a chain of typed steps that carries one message from raw input to delivered output.
Pipeline shape
Section titled “Pipeline shape”raw message → Deserialize → Extract → Validate → Transform → Serialize → SendEach step receives:
- a typed input,
- the pipeline context,
- a cancellation token.
Each step returns:
- a result,
- the updated context.
The transactional tutorial builds this pipeline one step at a time.
Step types
Section titled “Step types”| Step type | Use for | Failure means |
|---|---|---|
BusinessStep |
Validation and business rules. | A process owner must act. |
TechnicalStep |
Transformation, serialization, infrastructure calls. | The integration or platform failed. |
Step |
Rare steps that choose the result type directly. | The step decides. |
Finalizer |
Cleanup and reporting after the outcome is known. | It reacts; it does not advance the message. |
Result variants
Section titled “Result variants”| Result | Pipeline behaviour |
|---|---|
Success |
Run the next step. |
Cancelled |
Stop this message without error. Used by idempotency. |
BusinessFailure |
Stop this message and route a business incident. |
TechnicalFailure |
Stop this attempt. The runtime retries or dead-letters. |
A non-success result skips the remaining business steps. Finalizers still run.
Context
Section titled “Context”The context carries values that later steps need but should not recompute: message id, business subject, timestamps, correlation values. In the tutorial, the Deserializer reads the clock once and stores ProcessedAt; later steps stay pure and easy to unit-test.
State that must survive a run belongs outside the Component, usually in platform services such as idempotency and business incidents.