How to Prompt a Google Cloud PCA Scenario Coach

PCA coaches fail when they paste a remembered stem and mint a docs URL. Tag multi-region HA. Leave DOCS UNKNOWN.
The matching generator is the Google Cloud PCA Professional Cloud Architect Scenario Coach (Verify Against Exam Guide, No Dump) (Google Cloud Pca Professional Cloud Architect Scenario Coach No Dump) prompt. Browse related cards in the PromptDig library (Browse more prompts). When a filled run survives, share the version you actually use (Share a prompt).
Banner that this is not an exam dump
Coach Professional Cloud Architect scenarios from pasted exam-guide bullets. Verify against the guide. Not an exam dump. Start by filling Inputs, not by asking the model to remember last week's run. If a field is blank, write NONE or NOT IN INPUTS and leave it blank through Generate. The card is built so the model cannot honestly invent a number, owner, URL, or command that you did not paste.
Paste these fields before you hit run:
Exam-guide bullets I lock: [Guide]
Domain label I may use (or UNKNOWN): [Domain]
Scenario count I lock: [Count]
Words I must not use: [Banned]
What I must never invent (remembered stems, cut scores, docs URLs): [Never]
Output format: [Format]
Language: [Lang]
Docs URL I lock (or UNKNOWN): [Docs]
Audience I may name: [Audience]
That inventory is the honesty ledger. Anything that does not appear there is forbidden in the draft. If you catch yourself adding a nice-to-have after the run, you are no longer using the card. You are ghostwriting. Put the extra fact in Inputs and run again.
Tag every scenario with a pasted guide bullet
Generate is numbered on purpose. Do not skip a step because the first paragraph looked done. The early steps exist to stop later prose from smuggling claims.
Walk the Generate list in order:
- Honesty ledger: Guide nouns, Domain, Count, Docs, Lang, Audience. Forbidden: remembered stems, cut scores, invented docs URLs.
- Banner: not an exam dump. Unofficial practice from pasted guide bullets only.
- Scenarios: Count items. Each tags a Guide bullet. Tradeoff discussion from Guide nouns only.
- Verify: quote Docs or write DOCS UNKNOWN. Do not mint cloud.google.com paths.
- Refuse list: SAA-C03 stems, AZ-104 items, live-form claims, passing-score integers.
- Domain: quote Domain or write DOMAIN UNKNOWN.
- Never: do not invent a GKE node pool size. Do not claim an item appeared on a live form.
- Compliance pass: quote Banned and Never hits. Cut them. Count scenarios vs Count. Format as Format.
If a step asks for a version lock, quote the version from Inputs in the output. If a step asks for a refuse list, keep the refuse list in the published artifact, not in a sidebar you delete. Reviewers should see what the model was not allowed to do.
Verify against docs or write UNKNOWN
Most failures are the same shape: a missing field gets a confident fill. A conversion rate appears. A Gradle task appears. A flash point appears. A caption appears on a job that asked for slide text only. Your review is to search the draft for numbers, names, and commands, then grep Inputs. No match means cut.
Honor the constraints as hard stops, not vibes:
- PCA scenarios from Guide only. Not a dump and not SAA-C03.
- Never invent remembered stems, cut scores, or docs URLs.
- Stay at Count scenarios.
- No emojis.
When the card says not legal advice, not certification, not an exam dump, or not a caption engine, that sentence belongs at the top of the output. Deleting it to look more finished is how you inherit risk.
Refuse SAA-C03 and AZ-104 stems
Finish with the compliance pass the prompt already asks for. Quote the banned-word hits. Cut them. Print character counts when the job has a cap. Print word counts when the job has a budget. List gaps as gaps. Five missing facts are more useful than one smooth paragraph.
Tags on the card (pca professional cloud architect, exam guide scenario coach, no exam dump) are a reminder of the job shape, not an invitation to wander into a neighboring cluster. If you need a different surface, open a different PromptDig card rather than stretching this one.
Fill the card, then run
Replace every bracket. Run on ChatGPT, Claude, or Gemini. Read the ledger first, then the artifact. If the model invents a commit, KPI, DOI, PEL, bid, or logo, discard the run. Tighten Inputs. Run again. Share the filled card that survived, not the first draft that sounded done.