chrysopedia/backend/pipeline/quality
jlightner d75ec80c98 optimize: Stage 5 synthesis prompt — round 0 winner (0.95→1.0 composite)
Applied first optimization result: tighter voice preservation instructions,
improved section flow guidance, trimmed redundant metadata instructions.
13382→11123 chars (-17%).
2026-04-01 10:15:24 +00:00
..
fixtures feat: Generalized OptimizationLoop to stages 2-5 with per-stage fixture… 2026-04-01 09:24:42 +00:00
results optimize: Stage 5 synthesis prompt — round 0 winner (0.95→1.0 composite) 2026-04-01 10:15:24 +00:00
__init__.py feat: Created PromptVariantGenerator (LLM-powered prompt mutation) and… 2026-04-01 09:08:01 +00:00
__main__.py optimize: Stage 5 synthesis prompt — round 0 winner (0.95→1.0 composite) 2026-04-01 10:15:24 +00:00
fitness.py test: Built pipeline.quality package with FitnessRunner (9 tests, 4 cat… 2026-04-01 08:45:05 +00:00
optimizer.py optimize: Stage 5 synthesis prompt — round 0 winner (0.95→1.0 composite) 2026-04-01 10:15:24 +00:00
scorer.py feat: Generalized OptimizationLoop to stages 2-5 with per-stage fixture… 2026-04-01 09:24:42 +00:00
variant_generator.py feat: Added STAGE_CONFIGS registry (stages 2-5) with per-stage rubrics,… 2026-04-01 09:20:24 +00:00
voice_dial.py feat: Added VoiceDial class with 3-band prompt modification and ScoreRu… 2026-04-01 08:57:07 +00:00