Learn Agentic AI — Design · Build · Evaluate

Worksheet name

Progress

0 / 17 · 0%

Before you build

Agent Canvas filled and reviewed — All eight cells, with an explicit NOT for list. CanvasNotes
REMIT worksheet completed — Named owner, envelope in code, monitoring plan, identity, trust level. WorksheetNotes
System Prompt Builder run — Six ingredients in order; no single section longer than ~200 words. BuilderNotes
NIST GenAI Profile mapped — Which of the twelve risks apply, with explicit mitigations for each.Notes
EU AI Act classification — Documented (prohibited / high / limited / minimal / GPAI; provider vs deployer).Notes

Golden dataset of 20+ cases — Covering happy path, edge, adversarial, ambiguous, and handoff.Notes
Five tests passed — Happy, Edge, Adversarial, Ambiguous, Handoff. Five testsNotes
Red-team suite run — OWASP LLM Top 10 + your domain-specific attacks.Notes
Circuit breakers wired — Spend caps, tool-call caps, canary failures trigger automatic halt.Notes
Human oversight model chosen — Based on the risk × complexity matrix, and documented.Notes
Monitoring dashboard live — Action logs, tool-call regressions, quality regressions, cost, and latency.Notes
Kill-switch path — Exists outside the agent's runtime, and is known to on-call.Notes

Per-deploy regression run — Full golden dataset executed on every deploy; alert on any regression.Notes
Daily drift checks — Input distribution, output distribution, tool-call mix.Notes
Weekly human review — Sampled traces read, patterns noted, rubric updated.Notes
Authority review cadence — Monthly for new agents; quarterly for mature ones. Evidence-based promotion or demotion.Notes
Incident response plan — If a circuit breaker fires: who gets paged, what they do, when the board finds out.Notes