Discovery Policy

discovery controls selector discovery behavior.

discovery:
  max_concurrent: 5
  mode: auto
  lesson_cache: true
  replay_verify_threshold: 1.0
  static_mode_warning: true

Fields

Field	Type	Default	Description
max_concurrent	integer	5	Maximum concurrent discovery tasks.
mode	enum	auto	auto, static, or mcp.
lesson_cache	boolean	true	Use learned discovery lessons when available.
replay_verify_threshold	number	1.0	Required replay verification confidence, from 0.0 to 1.0.
static_mode_warning	boolean	true	Warn when static mode may miss dynamic behavior.

Modes

Mode	Meaning
auto	Let Yosoi choose the available discovery path.
static	Use static discovery behavior. Useful for deterministic constrained runs.
mcp	Use MCP-backed discovery when configured.