Skip to main content
INFORMATIVEDRAFTDocumentation Governance

SA-FLOW-02: SA Multi-Step Evaluation

Scope

This evaluation scenario validates multi-step execution in SA profile:

  • SA loads Context with multi-step Plan
  • SA processes steps by order_index sequence
  • Each step status updated correctly
  • Context remains stable across execution

Non-Goals

This scenario does NOT evaluate:

  • Single-step execution (see SA-FLOW-01)
  • Tool integration (see FLOW-03)
  • LLM enrichment (see FLOW-04)

Modules Involved

ModuleRole in Flow
ContextProvides stable execution boundary
PlanDefines multiple ordered steps

Evidence

TypeLocationStatus
Golden Flowtests/golden/flows/sa-flow-02-step-evaluation/Passed
Input Fixturestests/golden/flows/sa-flow-02-step-evaluation/input/Available
Expected Fixturestests/golden/flows/sa-flow-02-step-evaluation/expected/Available

Expected Behavior

  • Steps execute in order_index sequence
  • Step status transitions: pending → in_progress → completed
  • Context immutability preserved
  • All step_ids are valid UUID v4

Source of Truth: tests/golden/flows/sa-flow-02-step-evaluation/README.md