Process user-acquired content as local-only derived artifacts #47

Open
opened 2026-07-31 01:25:51 -04:00 by JackFrostbyte · 0 comments

Goal

Run validated user-acquired inputs through Arkive's existing processing adapters
without weakening bundle eligibility or duplicating the pipeline.

Requirements

  • The processing entry point receives an explicit core or local_overlay context.
  • Local inputs produce local-only document, chunk, and processing-manifest records.
  • Distribution scope is inherited and cannot be overridden by a CLI convenience flag.
  • Source, document, chunk, checksum, lineage, and adapter-version validation remains
    equivalent to the bundled pipeline.
  • Local failures, logs, and temporary files remain inside approved local paths.
  • Reprocessing is deterministic when the source bytes and tool versions are unchanged.
  • Recipe withdrawal prevents new acquisitions but does not falsify existing provenance;
    the UI/CLI marks affected local records accordingly.
  • A local artifact cannot satisfy or replace a bundled-core fixture.

Deliverables

  • Processing-context contract.
  • Local-overlay output layout.
  • CLI integration with the Issue #16 processing command.
  • Validation and reproducibility tests.
  • Documentation for deletion and reprocessing.

Acceptance criteria

  • One processing implementation serves both domains.
  • Every local derived artifact is traceable to a local acquisition record.
  • Local-only scope propagates through all derived records.
  • Core release/build inputs reject local artifacts.
  • Tests prove equivalent lineage and checksum guarantees across both domains.

Out of scope

  • Local index generation.
  • Search federation.
## Goal Run validated user-acquired inputs through Arkive's existing processing adapters without weakening bundle eligibility or duplicating the pipeline. ## Requirements - The processing entry point receives an explicit `core` or `local_overlay` context. - Local inputs produce local-only document, chunk, and processing-manifest records. - Distribution scope is inherited and cannot be overridden by a CLI convenience flag. - Source, document, chunk, checksum, lineage, and adapter-version validation remains equivalent to the bundled pipeline. - Local failures, logs, and temporary files remain inside approved local paths. - Reprocessing is deterministic when the source bytes and tool versions are unchanged. - Recipe withdrawal prevents new acquisitions but does not falsify existing provenance; the UI/CLI marks affected local records accordingly. - A local artifact cannot satisfy or replace a bundled-core fixture. ## Deliverables - Processing-context contract. - Local-overlay output layout. - CLI integration with the Issue #16 processing command. - Validation and reproducibility tests. - Documentation for deletion and reprocessing. ## Acceptance criteria - [ ] One processing implementation serves both domains. - [ ] Every local derived artifact is traceable to a local acquisition record. - [ ] Local-only scope propagates through all derived records. - [ ] Core release/build inputs reject local artifacts. - [ ] Tests prove equivalent lineage and checksum guarantees across both domains. ## Out of scope - Local index generation. - Search federation.
Sign in to join this conversation.
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Reference
Arkive/arkive#47
No description provided.