Module context_column 

Context Column — the cortical column abstraction for data source pipelines.

Each data source (filesystem, GitHub, Jira, DB, shell) is modeled as a neocortical column with four processing layers:

L4 (Input) — raw data ingestion, normalization → ContentChunks L2/3 (Predict) — compression mode selection, predictive coding L5 (Output) — verification, budget check, quality gate L6 (Feedback) — top-down modulation from active task context

Scientific basis: Mountcastle (Nature Rev Neurosci 2022) — every cortical column applies the same computational template to different input modalities.

The trait is async-ready (returns Results) so that network-backed columns (GitHub API, DB queries) work naturally alongside local columns (filesystem).

Structs

ColumnCompressed

Result of L2/3 (prediction/compression layer) processing.

ColumnContext

Parameters flowing top-down from L6 to modulate processing.

ColumnInput

Result of L4 (input layer) processing.

ColumnOutput

Result of L5 (output/verification layer) processing.

CrossSourceHint

A lateral connection hint to related data in other columns.

FilesystemColumn

The filesystem column — processes local files through the cortical pipeline.

ProviderColumn

Provider-backed column — wraps any ContextProvider as a cortical column.

Traits

ContextColumn

The cortical column trait — uniform processing pipeline for any data source.