Lesson 056: Evaluation harness depth
Focus
Treat placeholders as compulsory—swap nouns immediately after reading. Token Evaluation harness depth:56 keeps neighbouring lessons differentiable.
Key ideas
- Thread: Evaluation harness depth · drill v6 · spin
111278. - Habit: pair every model utterance with a trace_id you could paste into Grafana.
- Guardrail: write one RACI bullet referencing this lesson tomorrow.
Deep dive notebook
Synthetic drill artefacts
Logging field note
| Field | Retention |
|-------|-----------|
| trace_id | 21 days |
| prompt_hash_sha256 | permanent |
| completion_excerpt_redacted | 24h hot, then cold vault |
Heuristic `pii-mask-v1` tags sensitive spans before persistence.
Practice
Practice Draft three eval assertions QA must greenlight before launch. — 56 Bump literals mindset by 35.