John Lightner b16454994e MAESTRO: Implement Celery tasks (execute_run, execute_sweep) with synchronous fallback for single-container mode ... Created engine/tasks.py with: - execute_run and execute_sweep Celery tasks registered via autodiscover - SyncTaskResult class mimicking Celery AsyncResult for in-process mode - dispatch_run/dispatch_sweep helpers that route to Celery or sync based on config - Proper async-to-sync bridging for the async engine functions - 17 tests covering task execution, sync fallback, error handling, and Celery dispatch		2026-04-07 03:08:41 -05:00
..
adapters	MAESTRO: Implement OpenAI-compatible LLM adapter with streaming, retries, and tests	2026-04-07 02:35:52 -05:00
scorers	MAESTRO: Implement LLMJudgeScorer with configurable judge prompt, rating parsing, and response caching	2026-04-07 03:05:00 -05:00
__init__.py	MAESTRO: Create full directory structure with placeholder files	2026-04-07 01:40:27 -05:00
cache.py	MAESTRO: Implement ResponseCache layer with SHA-256 config hashing and hit-rate tracking	2026-04-07 02:37:58 -05:00
runner.py	MAESTRO: Implement run execution engine with Jinja2 templating, caching, scoring, and event bus	2026-04-07 02:48:20 -05:00
sweep.py	MAESTRO: Implement sweep orchestration engine with grid, random, and guided sweep types	2026-04-07 02:53:30 -05:00
tasks.py	MAESTRO: Implement Celery tasks (execute_run, execute_sweep) with synchronous fallback for single-container mode	2026-04-07 03:08:41 -05:00