The language layer has three canonical example families.
Confidence-gated classification
Demonstrates:
probable<T>- Confidence gates
ctx
The workflow classifies a message, then returns the label only when confidence clears the gate.
Semantic support routing
Demonstrates:
match similarwithin budget- Fallback control flow
- Context budgeting
The workflow routes customer messages by meaning and falls back when budget limits are exceeded.
Parallel research
Demonstrates:
agentspawngather@toolmemory- Budget fallback
The workflow launches one researcher per topic, gathers findings, and degrades to a simpler summary if needed.
Success criterion
For the language layer, the examples should pass voss check, compile to readable Python, and run with hermetic defaults where possible.
Do not use these examples to imply Voss is a Python replacement. They exist to validate AI workflow control.
