Skip to content

Artifact Mapping

Artifact mapping explains how ORCA Framework artifacts can be translated into other systems without pretending every concept maps perfectly.

What Mapping Is For

Use mappings when ORCA Framework artifacts need to move into:

  • harness-native formats
  • issue trackers
  • eval or reporting tools
  • future runtime integrations

Mapping Rules

  • keep the source artifact identity
  • preserve schema version
  • record lossy transformations
  • do not invent missing data to satisfy a target system
  • prefer explicit field maps over prose guesses

Typical Mapping Targets

Harness-Native Formats

Examples:

  • goal contract into a host-native goal request
  • runtime status into a host diagnostics panel or local status command
  • approval request into a harness-native approval prompt

Issue Trackers

Examples:

  • spec into an issue comment or linked document
  • regression task into an issue or checklist item
  • receipt into a work-item summary comment

Eval And Report Tools

Examples:

  • trace metadata into an eval runner
  • receipt summary into a benchmark or rollout report
  • QA finding into a regression candidate feed

Mapping Reality

Some fields map cleanly:

  • identifiers
  • status
  • timestamps
  • links
  • structured decision fields

Some fields do not:

  • nuanced rationale
  • freeform review notes
  • host-specific caveats
  • soft workflow interpretations

For those cases, keep a machine-readable core and preserve the richer Markdown artifact alongside it.

Mapping Artifact

Use templates/artifact-mapping.md to describe:

  • source artifact family and version
  • target system
  • field correspondence
  • known lossy areas
  • reviewer notes