Lesson 158: Benchmark scepticism rituals
Focus
Anchor this drill to one production LLM workflow—even hypothetical. Token Benchmark scepticism rituals:158 keeps neighbouring lessons differentiable.
Key ideas
- Thread: Benchmark scepticism rituals · drill v8 · spin
760309. - Habit: attach a trace_id to every completion you would paste into an ops dashboard.
- Guardrail: add one RACI bullet for prompt or index changes before tomorrow's standup.
Deep dive notebook
Synthetic drill artefacts
Red-team tableau
1. Actor **support_agent**
2. Injection `INJECT-183`
3. Detector `policy-tier-3` → pager `llmops-oncall-2`
4. Log fields `trace_id,tenant_id,redaction_tier`
Practice
Practice Simulate degraded retrieval once; capture user-facing fallback copy. — 158 Bump 25.