Use the practice organisation and source records from lesson 1.
All practice records are synthetic. Work manually or with an optional AI assistant; no paid product, external account or connected inbox is needed.
Open the source pack · Download the workflow templates
Your assignment
Using P1–P2, E1–E7, C1–C2 and M1, prepare a handoff packet for Morgan. Do not send anything. E7 supersedes Riley's earlier availability, so recheck scheduling instead of reusing the previous answer.
Include:
- A briefing of at most 180 words covering the immediate review, supplier dependency, launch uncertainty and scheduling problem, with source IDs.
- An action queue with owners, due times and unknowns.
- A draft to Riley explaining that the revised window does not currently fit and asking for alternative availability. Keep private calendar event names out of the external draft.
- A separate approval queue for external communication, budget disclosure and any purchase decision.
- A short test log: two errors you looked for, what you found and what you changed.
Check your answer — open after your attempt
With E7, Riley is available only 15:00–16:00. Morgan needs a gap before 15:30 coaching, so 15:00–15:30 cannot work. After coaching ends at 16:00, the gap moves the next possible start to 16:15, outside Riley's window. There is no valid 30-minute slot under the stated constraints. Do not silently shorten the meeting, drop Sam or remove Morgan's buffer.
Example Riley draft: “Thanks for the update. The revised Tuesday window does not fit a 30-minute meeting with Morgan and Sam. Could you share another time window? We’ll check it and send a proposal for confirmation.” Label it a draft awaiting Morgan's review.
Example briefing: “Morgan's comments on the launch outline are needed before 11:00 Monday; this is not a launch-approval request (E1). Sam will collect supplier delivery and cancellation details for Jules by Tuesday 16:00; the quote expires Wednesday 12:00 and purchase approval is still pending (M1, E3). Sam's readiness checklist is due Wednesday 15:00. The launch date remains undecided and Riley's feedback is pending (M1). Riley's revised Tuesday availability leaves no valid 30-minute slot under the current calendar constraints (E7, C1, C2, P2). Please review the alternative-availability draft and decide whether any approved budget information may be shared (P1, E6).”
Rubric for self assessment
Score each area 0, 1 or 2: 0 = absent or materially wrong; 1 = present with a correctable gap; 2 = complete and supported.
- Evidence: key claims have correct source IDs.
- Accuracy: deadlines, owners, decisions and uncertainty are preserved.
- Scheduling: the E7 conflict is correctly identified without changing constraints.
- Discretion: no private budget or unnecessary calendar detail appears in the external draft.
- Approval: drafts, disclosure and purchase decisions remain visibly pending.
- Usability: a reviewer can see the next action and the unresolved question quickly.
Suggested practice target: at least 10/12 with no zero in accuracy, discretion or approval. This is an instructional target, not an accredited qualification or predictor of job performance. Revise and explain each correction. Completing the course does not guarantee employment or time savings.
Keep your portfolio honest
Label the packet “Synthetic executive-assistant workflow exercise.” State which parts you created manually and where AI helped, if it did. Include one before-and-after correction showing your review. Do not describe the fictional organisation as a client or claim these were real executive results.
Sources and teaching notes
[1] NIST, Generative Artificial Intelligence Profile, AI 600-1: https://nvlpubs.nist.gov/nistpubs/ai/NIST.AI.600-1.pdf — supports the narrow factual warning that generative AI can confidently produce false content. The workflows, records, answers and rubric here are original teaching material, not NIST certification or endorsement.
[2] OWASP, Top 10 for Large Language Model Applications: https://owasp.org/projects/top-10-for-large-language-model-applications — background on application-level LLM risk. The course's review and access boundaries are instructional safeguards; a prompt alone is not a security control.
Sources checked 12 October 2026. No product-specific interface, paid feature or automatic action is required.