Skip to content

Capability Map

The capability map for bijux-canon-agent should let a reviewer connect workflow promises to the code that coordinates roles and emits traces. If orchestration behavior cannot be mapped clearly, the package starts to look magical.

Capability Flow

flowchart LR
    workflow["workflow promises"]
    modules["application, role modules, services, interfaces, and artifacts"]
    outputs["trace-bearing orchestration output"]

    workflow --> modules --> outputs

This page should show orchestration capability as something a reviewer can locate, not something they have to trust. Workflow claims need a visible path through modules and traces.

Capability To Code

  • application/ and orchestration flows own role sequencing and workflow coordination
  • agent services and role modules own the behavior of local actors inside a workflow
  • interfaces/ and artifacts own the surfaces where callers and operators inspect workflow behavior

Visible Outputs

  • trace-bearing workflow execution records
  • role-specific outputs that remain attributable to a step and an order
  • agent-facing contracts that expose orchestration intentionally

Design Pressure

Agent capability starts to look magical when role coordination and trace output cannot be tied back to named code areas. The package has to keep workflow promises grounded in visible modules and artifacts.