--- title: "CCAO-F Worked Scenarios: 10 Items Explained | Cred Farmer" url: https://credfarmer.com/blog/ccao-f-practice-questions description: "Ten CCAO-F worked scenarios written to Exam Guide v1.0: original business items across all seven domains, the key, and why every wrong option fails." last_updated: 2026-10-03 publisher: "Cred Farmer, an independent study platform, not affiliated with or endorsed by Anthropic or Pearson VUE" --- Claude Certification Blog # CCAO-F worked scenarios: ten practice items across the seven Associate domains CCAO-F worked scenarios are worth the time only when every option is explained. Here are ten original ones across all seven Associate domains, weighted like the blueprint, each with its key and a reason every distractor fails. 10 worked items · Single and select-two · 7 official domains Published 17 September 2026Updated 3 October 202619 min read By the Cred Farmer editorial team.Checked against the official sources on 3 October 2026. These ten CCAO-F worked scenarios are original Cred Farmer items written to the CCAO-F Exam Guide v1.0 (July 2026): a short workplace scenario with a constraint, then four options, or five where an item asks you to select two. They cover all seven official domains in rough proportion to weight, every wrong option carries a reason it fails, and none is drawn from a live exam form or from the question bank behind the login. Answer each one cold, then score yourself on the miss map. The [official CCAO-F exam guide](https://anthropic-partners.skilljar.com/page/partner-certifications) publishes three CCAO-F sample questions of its own in Section 8 and says they are not drawn from the live item bank. The [Claude certification sample questions](https://credfarmer.com/blog/claude-certification-sample-questions) hub works one item per code; this post is its Associate spoke. The blueprint and fee are on the [Claude Certified Associate - Foundations](https://credfarmer.com/certifications/claude-certified-associate-foundations) page, and the [guide to Claude certification practice exams](https://credfarmer.com/guides/claude-certification-practice-exams) covers what official practice material exists. The ten are spread by official weight, so output evaluation at 21 percent gets two items and troubleshooting at 10 percent gets one. One of the ten asks for two answers. Read the stated constraint before the options, decide what it rules out, then compare what is left. **Figure:** Seven CCAO-F domains by official weight, from 10 to 21 percent, with one or two items from this set in each - Official weight · items here - D1 Prompting and Task Execution · 14% · 1 item - D2 Output Evaluation and Validation · 21% · 2 items - D3 Product and Model Selection · 12% · 1 item - D4 Workflow Integration and Solution Design · 16% · 2 items - D5 Configuration and Knowledge Management · 12% · 1 item - D6 Governance, Risk, and Responsible Use · 15% · 2 items - D7 Troubleshooting and Optimization · 10% · 1 item Weights from the CCAO-F Exam Guide v1.0, Section 6; the item counts are this post's ten. ## How do you match the prompt to the task in front of you? Domain 1, Prompting and Task Execution, is 14 percent of the blueprint, and its items rarely ask you to write a prompt. They ask which kind of task a team actually has, then which move fits: decomposition for a multi-part request, iteration for a near miss, a different strategy when the task type was misread. The item below turns on that last case. Question 1 · Written for this post · Domain 1 · single response Asked for the one best campaign idea with evidence, Claude gave one concept and three survey figures the team never collected. The team needs many options for a workshop. What prompt change fits? - **A.** Keep one idea but source each figure - **B.** Ask for three concepts, each with projected revenue - **C.** Use research mode to find market figures before choosing - **D.** Ask for fifteen distinct ideas across set angles, with no statistics **Reveal the answer and the reason for every option** Answer: D. Choose the prompting strategy for the task type. **A.** This keeps the single answer the workshop does not want. Sourcing figures for one concept still gives the team nothing to choose between. **B.** Projected uplifts invite more invented numbers, and three concepts is still a narrow field for a workshop. **C.** Research suits a decision that needs evidence. The team first needs a wide range of ideas, which is a brainstorming task. **D.** Correct. The task is brainstorming, so breadth matters and evidence is not yet needed. Dropping the request for evidence also removes the prompt that invited invented figures. Adapt the prompting strategy to the task type. A workshop that needs options calls for brainstorming: many distinct ideas across stated angles, with judgment held back. Asking for one best idea with evidence turned it into an analysis task, and the missing research produced invented statistics. Sourcing, projections and research mode all keep the wrong task type in place. ## Which output can you trust, and what shape should it take? Output Evaluation and Validation is the heaviest domain at 21 percent, so two of the ten sit here. The first tests what an internal contradiction tells you about the process behind a document, and what must happen before anyone edits it. The second is a select-two on output format: two audiences with two next actions need two containers. In both, the tempting wrong answers make the output tidier, not truer. Question 2 · Written for this post · Domain 2 · single response In Claude's supplier risk summary, the table says 7 of 22 suppliers are single-sourced, the text says nearly half, and one supplier is listed under two countries. What should you do first? - **A.** Check the count and each country against the source data - **B.** Ask Claude to reconcile the table and text and use its answer - **C.** Edit the text so it matches the table - **D.** Mark the figures as indicative and send it on **Reveal the answer and the reason for every option** Answer: A. Inconsistency is evidence about the facts. **A.** Correct. Two figures for one count and two countries for one supplier show the facts are unreliable. The source data settles both before anything is tidied. **B.** The same process produced both versions. Claude can pick one confidently without checking the source, so its answer is not verification. **C.** Tables are not more reliable than prose by default. The table's count could be the wrong one. **D.** A caveat hides a known inconsistency on a material fact. The paper should not go out with figures known to conflict. Inconsistencies inside an output are a sign that some of its facts are wrong. A count stated two ways and a supplier placed in two countries should be checked against the source data before the paper is finalized. Asking the model to choose, trusting the table by default or adding a caveat leaves the error unresolved. Question 3 · Written for this post · Domain 2 · select two Claude listed 40 reconciliation exceptions. The payments team imports rows with fixed field names into its tracking system, and an executive wants a two-minute read before a morning call. Which two outputs fit? (Select two.) - **A.** A CSV of all 40 exceptions with field names matching the import - **B.** One long chat reply listing every exception in prose - **C.** One slide artifact meant to serve both audiences - **D.** The full 40-row table for the executive as well - **E.** A short inline summary of total exposure and the largest items **Reveal the answer and the reason for every option** Answer: A and E. Format follows how each audience will use it. **A.** Correct. The tracking system consumes rows with fixed fields, so structured data matching its import format is the right output for the payments team. **B.** Prose cannot be imported, and forty items in prose is too long for a two-minute read. It fails both audiences. **C.** A slide cannot be loaded into the tracking system, and one output cannot fit two very different uses. **D.** Forty rows is not a two-minute read. The executive needs the total and the largest items, not every exception. **E.** Correct. The executive needs something readable in two minutes, so a brief inline summary of what matters fits that use. Selecting an output format means asking how each audience will use the result. A system that imports rows needs structured data with fixed field names, while a person with two minutes needs a short inline summary. One format for both, whether prose, slides or a full table, serves neither use well. ## When has one long conversation stopped being the right container? Product and Model Selection includes an objective on context limits and when to restart, summarise or persist. The signature is a session that re-asks settled questions or contradicts agreed decisions. The item below adds two constraints that separate the plausible remedies: how long the work will run, and who else has to pick it up. Weigh every option against both. Question 4 · Written for this post · Domain 3 · single response After three hours in one chat building a due diligence checklist, Claude re-asks settled questions. The matter runs six more weeks, and colleagues will pick up the work. What should you do next? - **A.** Stay in this chat and paste back each decision it forgets - **B.** Verify a decision summary and keep it in a Project - **C.** Switch this chat to the strongest model to hold the history - **D.** Start a new chat with just the draft **Reveal the answer and the reason for every option** Answer: B. Summarize, verify, then persist. **A.** The overloaded context stays, and pasting decisions back adds to it. The drift will continue, and colleagues cannot use this chat. **B.** Correct. A checked summary carries the settled decisions forward, and a Project keeps it with the current draft for every new conversation, including colleagues' work. **C.** The problem is context load, not model strength. Six more weeks in the same conversation will only make it worse. **D.** The draft does not record why things were settled or which items were excluded. Those decisions would be argued again. Managing context limitations means knowing when to summarize, restart and persist. Re-asked questions show the conversation has outgrown what Claude handles well. A verified summary of decisions and open questions, kept in a Project with the current draft, restores reliable context for the weeks ahead and gives colleagues the same starting point. Staying put, switching models or restarting without the decisions does not. ## Where does Claude fit in a workflow, and how do you tell leadership? Workflow Integration and Solution Design is 16 percent of the exam. Its items give a team a mix of tasks with different stakes, then ask what an associate should propose. The first item below turns on segmenting a use case by risk and by who makes each decision. The second is about communicating value: the benefit and the control that makes it safe belong in the same sentence, and a promise the pilot never measured belongs nowhere. Question 5 · Written for this post · Domain 4 · single response Complaints split into information requests (half), refund claims with legal deadlines (30 percent) and service incidents, some about safety (20 percent). You must propose where Claude fits. Which analysis is best? - **A.** Automate replies for all types and sample a few for review - **B.** Limit Claude to information requests and leave the rest alone - **C.** Segment the types by risk and who decides - **D.** Launch one prompt for all and see what breaks **Reveal the answer and the reason for every option** Answer: C. Analyze the use case by segment. **A.** Sampled review is too weak for refunds with legal deadlines and safety incidents. One approach for all ignores the differences in risk. **B.** Refund cases can still be prepared by Claude for staff to decide. Ruling out every other type gives up safe gains. **C.** Correct. Each type has a different consequence and decision owner. Claude can draft information replies, prepare refund cases for staff and stay out of safety decisions. **D.** Learning by failure is not acceptable where refunds have legal deadlines and some incidents involve safety. Analyzing a use case well means breaking it into segments with different structures, consequences and decision owners. Information requests suit drafting for agent review, refund claims can be prepared for a person to approve within the legal timescale, and safety incidents stay with staff. Treating all complaints the same, or excluding everything but one type, misses where Claude can and cannot help. Question 6 · Written for this post · Domain 4 · single response A pilot cut drafting time for enrollment emails from 25 to 8 minutes, and 12 percent of drafts needed a fee or deadline fix. What should you tell leadership? - **A.** The time saved, the error rate and the review that catches errors - **B.** The saving as next year's headcount cut - **C.** That uploading fee schedules will bring errors to zero - **D.** That staff were satisfied, which shows accuracy **Reveal the answer and the reason for every option** Answer: A. Communicate the benefit and the control together. **A.** Correct. Leadership needs the measured benefit, the measured error rate and the control that keeps errors from reaching students, together. **B.** The saving depends on staff reviewing every draft. Treating it as a headcount cut overstates the gain and removes the control. **C.** Better sources may reduce errors, but promising zero is a claim the pilot cannot support. **D.** Satisfaction is not accuracy. One draft in eight still needed a factual correction. Leadership should hear what the pilot gained, what went wrong and which control handles it, all in one place. The pilot saved real time and also produced fee and deadline errors that a review step caught. Turning the saving into headcount, promising zero errors or citing satisfaction as accuracy each misstates what the pilot showed. ## Which written rules keep a Claude deployment inside its remit? Configuration and Knowledge Management at 12 percent and Governance, Risk, and Responsible Use at 15 are separate domains whose items fail the same way: a rule that should exist in writing has been replaced by a hope. Item 7 is a Project instruction that exhorts instead of specifying. Items 8 and 9 move the rule outside the tool: who may make a consequential decision, and which route a new tool travels before anyone's data goes into it. Disclosure, emphasis and personal judgement are offered as substitutes, and none is one. Question 7 · Written for this post · Domain 5 · single response A support Project's only instruction says: "Be a helpful expert and never make mistakes." Replies vary in tone, promise refunds the team cannot approve, and quote prices not on the uploaded list. Which rewrite fits? - **A.** Add, in capitals, that every reply must be one hundred percent accurate - **B.** Set role and tone, name the price list, and forbid refund promises - **C.** Paste twenty good past replies into the instruction - **D.** Add a line asking for a warm, steady tone **Reveal the answer and the reason for every option** Answer: B. Name the sources, limits and tone. **A.** Emphasis is not information. Claude still would not know which source governs prices or which commitments are off limits. **B.** Correct. Each failure maps to a missing instruction: no stated tone, no named pricing source and no limit on commitments. This rewrite supplies all three. **C.** Examples hint at tone but state no rules. Pricing and refund limits would still be left to inference. **D.** This addresses only the tone. The refund promises and wrong prices come from missing rules on sources and commitments. Effective system-level instructions state the role and audience, name the sources Claude must use, set limits on what it may commit to and fix the tone. A vague demand for perfection gives Claude nothing to act on, which is why replies vary and stray. Emphatic wording, examples alone or a tone line each fix at most one of the three failures. Question 8 · Written for this post · Domain 6 · single response An HR director wants Claude to cut screening time by ranking 600 nursing applicants, shortlisting 80 and auto-rejecting the rest. Policy requires a documented human decision for every rejection. What is the most appropriate use? - **A.** Automate it as proposed, noting AI help in each rejection letter - **B.** Rank all 600 and let one recruiter approve the list in one step - **C.** Summaries against published criteria, with people deciding - **D.** Use Claude only for the rejection letters **Reveal the answer and the reason for every option** Answer: C. Keep consequential decisions with accountable people. **A.** Disclosure does not create the documented human decision the policy requires. Rejections would still be made by the tool. **B.** One approval of a ranked list is not a decision on each applicant. It turns human review into a formality. **C.** Correct. Summaries against the criteria save screening time, while shortlisting and every rejection stay with recruiters, as policy requires. Testing the summaries for bias on a sample adds a check. **D.** This is permitted, but it is not the most appropriate use: recruiters would still screen all 600 applications by hand, which is the time the director wants to cut. Identifying appropriate use means separating help with consequential decisions from making them. Hiring decisions affect people's livelihoods and are subject to anti-discrimination law, and policy requires a documented human decision for each rejection. Claude can summarize applications against published criteria to save time, while people shortlist and reject. Automated rejection and a one-click approval of a ranking both miss that balance, and limiting Claude to the letters leaves the screening time in place. Question 9 · Written for this post · Domain 6 · single response A dispatcher installed a "powered by Claude" browser extension that sends every open tab to an outside service. Policy lists approved tools and names an owner for new ones. What should happen? - **A.** Keep using it on tabs with no customer names - **B.** Keep using it if the vendor says nothing is stored - **C.** Trial it for another week, then report to the owner - **D.** Stop using it; submit it for approval **Reveal the answer and the reason for every option** Answer: D. Follow the approved list and its route. **A.** A tool built on Claude is not the same as the approved Claude workspace. Limiting which tabs are open does not make an unapproved tool approved. **B.** A vendor's assurance is not the firm's approval. The policy names who decides, and it is not the vendor. **C.** Using a tool first and asking afterward reverses the approval route. Customer data would keep flowing in the meantime. **D.** Correct. The extension is not on the approved list, so it stays unused until the named owner approves it through the route the policy sets. Following organizational AI policy means using only approved tools and routing new ones through the named approval process before use. An extension that sends every tab to an outside service is a new tool handling company data, however it describes itself. Narrowing its use, relying on the vendor's word or trialing first and reporting later all bypass the governance the policy sets. ## What do you check first when a Project that worked starts failing? Troubleshooting and Optimization is the smallest domain at 10 percent, and its first objective is diagnosis before repair: locate where the failure enters before changing the prompt, the model or the knowledge. The item below gives you a change and a symptom, and tests whether you connect them by evidence before touching anything. Question 10 · Written for this post · Domain 7 · single response Claim letters from a Project began quoting wrong deductibles once a new policy version was uploaded beside the old one. Both still apply until renewal. What should you do first? - **A.** Tell Claude to always check and use the correct deductible - **B.** Trace bad letters to their source and label each version by its dates - **C.** Remove the new version to restore the old setup - **D.** Move the Project to the strongest model available **Reveal the answer and the reason for every option** Answer: B. Diagnose the source before touching the prompt. **A.** Claude cannot tell which version applies when nothing says so. A general instruction to be correct adds no information. **B.** Correct. Tracing failing letters confirms which version each amount came from, and labeling the versions by date lets an instruction select the right one for each policyholder. **C.** Both versions are needed until renewal. Removing one makes letters wrong for the policyholders on the newer terms. **D.** The failure began with the upload, not the model. A stronger model faces the same unlabeled conflict between sources. Diagnosing poor outputs starts with what changed. The errors began when a second policy version was added, and both must stay, so the fix is to confirm the cause by tracing failing letters and then make the sources distinguishable so the right one can be chosen. Vague accuracy instructions, deleting a needed source or upgrading the model skips the diagnosis. ## Where did these CCAO-F worked scenarios catch you? A miss map by domain Score the ten, then count misses by domain rather than in total. Ten CCAO-F questions cannot produce a percentage that means anything against a scaled cut of 720, and the score report gives percent-correct by domain, so that is the unit to practise reading. Each row names the items here, the objective to reread if you missed one, and the free mock's untimed filter for that domain. | Domain | Items here | Your misses | Objective to reread | Untimed drill | | --- | --- | --- | --- | --- | | D1 Prompting and Task Execution (14%) | 1 | __ of 1 | O04: prompting strategy by task type | [Domain 1 untimed](https://credfarmer.com/practice/ccao-f?domain=1) | | D2 Output Evaluation and Validation (21%) | 2 and 3 | __ of 2 | O06: inconsistencies as a signal about the source; O10: output format for its next use | [Domain 2 untimed](https://credfarmer.com/practice/ccao-f?domain=2) | | D3 Product and Model Selection (12%) | 4 | __ of 1 | O14: context limits: restart, summarise or persist | [Domain 3 untimed](https://credfarmer.com/practice/ccao-f?domain=3) | | D4 Workflow Integration and Solution Design (16%) | 5 and 6 | __ of 2 | O15: requirements and use cases by segment; O18: value and limitations stated together | [Domain 4 untimed](https://credfarmer.com/practice/ccao-f?domain=4) | | D5 Configuration and Knowledge Management (12%) | 7 | __ of 1 | O21: system-level instructions: sources, limits, format | [Domain 5 untimed](https://credfarmer.com/practice/ccao-f?domain=5) | | D6 Governance, Risk, and Responsible Use (15%) | 8 and 9 | __ of 2 | O22: appropriate and inappropriate use cases; O24: the AI policy and its approval route | [Domain 6 untimed](https://credfarmer.com/practice/ccao-f?domain=6) | | D7 Troubleshooting and Optimization (10%) | 10 | __ of 1 | O26: diagnose before changing anything | [Domain 7 untimed](https://credfarmer.com/practice/ccao-f?domain=7) | **Figure:** Count misses by domain: none means move on, one means reread the objective, two means run the domain filter untimed - Score the ten items - then count misses by domain - 0 misses - Move on - The full mock rechecks it - 1 miss - Reread the objective - Then redo the item cold - 2 misses - Run the domain filter - Untimed, in the free mock A Cred Farmer reading of a ten-item score, not an official rule. Domains with one item here cap at one miss, so treat a miss there as a reread, not a verdict. Two misses in the same domain is the pattern worth acting on: run that domain's filter untimed, then sit the full timed form. The [two-week CCAO-F study plan](https://credfarmer.com/blog/two-week-study-plan-ccao-f) schedules the rereads around two timed sittings, the [CCAO-F cheat sheet](https://credfarmer.com/blog/ccao-f-cheat-sheet) compresses each domain to its key points, and [why exam dumps are worth so little](https://credfarmer.com/blog/claude-certification-exam-dumps) explains what a memorised answer letter costs. ### Ten items are a sample, not a score Two misses in the same domain is a pattern; one miss anywhere is noise until the full 60-item form confirms it. Nothing on this page is a scaled score, a pass verdict or a readiness figure, and it does not predict an official result. Question bank dated 2 October 2026. Checked against the CCAO-F Exam Guide v1.0 (July 2026) on 17 September 2026. Cred Farmer is not affiliated with Anthropic; the [CCAO-F](https://credfarmer.com/certifications/claude-certified-associate-foundations) page records what the official sources say and when they were last read. ## Key takeaways - **Read the constraint before the options.** Every item here turns on a stated fact: a policy, an upload, who picks up the work next. - **Count misses by domain, not in total.** Ten items give no percentage worth quoting against a scaled cut of 720. - **The tidy answer is often the trap.** Polishing, a disclaimer or a model switch each treats a symptom the scenario has already explained. - **A rule that is not written down is a hope.** Instructions, policies and who decides all fail the same way when they live in one head. - **Practice evidence is not a pass probability.** Only first sittings of the aligned forms set the band of a Cred Farmer readiness score, and it does not predict an official result. ## Sit the whole 60-item form under the clock The free CCAO-F mock is the full 60 items in 120 minutes, no account, with a reason for every option and a raw count by domain beside the official weight. Signed in, five blueprint-aligned timed forms add 300 more items; your first sitting of each sets the band of a Cred Farmer readiness score, which does not predict an official result. [Start the free CCAO-F practice exam](https://credfarmer.com/practice/ccao-f) Already scored the mock? [Sign in for the five timed forms](https://credfarmer.com/login?next=/c/ccao-f/practice). Questions ## Frequently asked Short answers to the questions this post leaves open. **Are these real CCAO-F exam questions?** No. All ten were written fresh for this post from the CCAO-F Exam Guide v1.0 and its published objectives; none comes from a live exam form or from the gated Cred Farmer bank. The guide's own three sample items are illustrative too, and it says they are not drawn from the live item bank. **How many questions are on the CCAO-F exam?** The CCAO-F has 60 items in 120 minutes, multiple-choice and multiple-response, and each item states how many responses to select. The result is a pass or fail with a scaled score from 100 to 1,000; the cut is 720. The fee is $99 and the credential is valid for 12 months. **Is there a free CCAO-F mock exam?** Yes. The free Cred Farmer mock at /practice/ccao-f runs the full 60-item form under a 120-minute clock with no account, then shows a raw count by domain and a reason for every option. It never reports a scaled score or a pass verdict, because no practice percentage converts to the 720 cut. **How should I score myself on these ten CCAO-F questions?** Count misses by domain, not in total, and treat a select-two item as a miss unless both picks match the key. One miss in a domain means reread the objective in the miss map and redo the item cold a day later; two means run that domain's untimed filter in the free mock first. **What score do I need on practice questions to pass the CCAO-F?** No practice percentage converts to the exam's scaled cut of 720. The measure we use is Cred Farmer's readiness score: signed in, it reaches Exam-ready (90) only when your first sittings of the five blueprint-aligned forms average 800 or more on our own practice scale, about 78% correct. The free items here never count. No real exam result has checked that scale, so it does not predict an official result. Keep reading ## Related posts Article ### Ten CCDV-F worked scenarios, explained option by option Ten CCDV-F worked scenarios written to Exam Guide v1.0: original stems across six domains by weight, the key, why every wrong option fails, and a miss map. Read the post → Article ### Ten CCAR-P architecture scenarios explained across the seven Professional domains Ten CCAR-P architecture scenarios explained against Exam Guide v1.0: original items across all seven domains, the key, and why every wrong option fails. Read the post → Guide ### Claude certification sample questions, explained Six original Claude certification sample questions with worked explanations, the official format, clearly labelled mock data, and a practical reading method. Read the post → Not affiliated with, or endorsed by, Anthropic or Pearson VUE. Details are summarised from publicly published program information and can change. Always confirm against the official exam guide before booking.