Schema enforcement already prevents LLM from returning fields outside the defined response_schema (_SingleStepOut, _BatchStepOut). Explicit field bans (dose, amount, quantity, application_amount) are redundant and add unnecessary token cost. Removed: - 'KRYTYCZNE' warning about schema violations - 'ZABRONIONE POLA' explicit field list - 4-line 'ABSOLUTNIE ZABRONIONE' dose prohibition section Token savings: ~80 tokens per prompt (system instruction overhead) Trust the schema - cleaner prompts, same enforcement. |
||
|---|---|---|
| .. | ||
| alembic | ||
| innercontext | ||
| tests | ||
| .env.example | ||
| .python-version | ||
| alembic.ini | ||
| db.py | ||
| main.py | ||
| pyproject.toml | ||
| README.md | ||
| skincare.yaml | ||
| test_query.py | ||
| uv.lock | ||
See the root README for setup and usage instructions.