- Wire resolveProfileDefaults into loadEffectiveGSDPreferences so
token_profile: budget actually sets cheaper models and skips phases
- Add /gsd rate <over|ok|under> command to submit user feedback on
model tier assignments, completing the adaptive routing feedback loop
- Document that models config is required for dynamic routing activation
- Document ceiling behavior when dynamic routing is active
- Document reassess_after_slice as required for reassessment
Closes#1505 (partial — escalateTier wiring deferred to follow-up)
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Wire semantic chunker into inlineFileSmart() for large file context selection
- Use inlineFileSmart for knowledge file in buildExecuteTaskPrompt (TF-IDF relevance)
- Add compression_strategy and context_selection preferences with profile defaults
- Add resolveCompressionStrategy() and resolveContextSelection() resolvers
- Add cacheHitRate and compressionSavings to UnitMetrics
- Add aggregateCacheHitRate() for session-wide cache performance
- Update token-optimization.md with compression, chunking, and distillation docs
- Add 12 integration tests for optimization preferences and modules
