Curriculum
The full CCAO-F curriculum: 7 domains, 30 lessons. Every lesson covers the concept, key points to know, and exactly what the exam tests about it.
Domain 1
14% of examPrompting and Task Execution
Write prompts that give Claude what a capable colleague would need, break big requests into checkable steps, improve results with specific feedback, and tune your approach to the kind of task.
Effective Business Prompts
A strong business prompt reads like a clear brief to a capable new colleague. It states the role Claude should take, the context behind the request, the constraints that matter, and the format the result should arrive in. Structure beats intensifiers: adding 'be thorough' helps far less than adding the missing context.
Task Decomposition
Compound requests fail quietly because Claude has to juggle too many goals in one pass. Splitting the work into a chain of smaller prompts, each with a checkable output, gives you a point to inspect and correct before errors carry forward.
Iterative Prompt Improvement
The first response is a draft to refine, not a verdict. Improvement comes from specific feedback that says what to keep, what to change, and why. Regenerating and hoping gives you a new random draw instead of progress.
Prompt Strategy by Task Type
Different tasks need different prompting. Analysis wants source material and a clear question. Research wants scope and a way to check claims. Drafting wants audience, voice, and examples. Brainstorming wants room to explore before you narrow.
Domain 2
21% of examOutput Evaluation and Validation
The heaviest domain. Judge whether outputs answer the question, spot hallucinations and bias, verify against a source of record, know when a human must review, and shape outputs for the audience.
Accuracy and Completeness
Before polishing, check two things: did the output answer the question that was asked, and did it leave out anything material? A fluent response that quietly skips a caveat, an exception, or half of the request is still a failed response.
Hallucinations, Inconsistencies, and Bias
Claude can produce text that sounds right but is not grounded in your source or in fact. Learn the red flags: invented numbers and citations, detail beyond what the source contains, internal contradictions, uniform confidence on every claim, and one-sided framing.
Fact-Checking and Validation
Validation means checking a claim against something independent of the model: the source of record. Self-reported confidence is not validation, and asking Claude to double-check its own answer is weaker than looking at the original document, system, or authority.
When Human Review Is Required
Some outputs should never leave your hands without a person reviewing them. The more external, consequential, or regulated the use, the stronger the need for human review. Claude drafts and assists; accountable people decide.
Edit and Adapt for Audience
A correct answer can still be wrong for the reader. Adapting tone, length, and level of detail for a specific audience is a normal part of using Claude, and comparing drafts side by side helps you choose deliberately.
Output Formats
The shape of an answer matters. Short conversational replies suit inline responses, standalone documents and reusable pieces suit Artifacts, and comparisons or lists suit tables and structured layouts. Pick the format that fits how the result will be used.
Domain 3
12% of examProduct and Model Selection
Choose the right claude.ai feature for the job, understand the Haiku, Sonnet, and Opus tiers, balance cost, speed, and quality, and work within context limits and memory.
Product Features
Claude.ai offers several surfaces, and each fits a different kind of work. Chat is for one-off conversation, Projects hold recurring work with shared instructions and knowledge, research mode gathers and synthesises sources, and Artifacts hold standalone deliverables.
Model Tiers
Claude comes in tiers with different characters. Haiku is the fast, lightweight option for simple high-volume work. Sonnet is the balanced everyday choice. Opus is the most capable tier for complex reasoning and nuanced judgement.
Cost, Speed, and Quality Alignment
Every task sits somewhere on a triangle of cost, speed, and quality. The skill is picking the smallest model and lightest approach that meets the quality bar, rather than defaulting to the strongest tool everywhere.
Context Limits and Memory
A conversation has a finite context window. Long threads degrade, details get lost, and earlier instructions can be diluted. Know when to restart, when to summarise, and when to persist information in a Project or memory instead of relying on the thread.
Domain 4
16% of examWorkflow Integration and Solution Design
Turn Claude into part of how work gets done: define the deliverable first, pick strong-fit use cases, pilot and expand, augment or redesign processes, and explain value and limits to stakeholders.
Requirements and Use Cases
Good solutions start before the first prompt. Define the deliverable, the audience, the sources, and what success looks like. A clear requirement turns a fuzzy idea like 'use AI for reports' into something Claude and your team can execute and measure.
Research, Planning, and Process Optimization
Claude is strong at some work and weak at other work. Use it for synthesis, drafting, structuring, and analysis of material you provide. Be cautious where the task needs live data, guaranteed precision, or authoritative decisions.
Solution Design and Iteration
Roll out in small steps. Pilot with a narrow use case, measure against the success criteria you defined, learn what breaks, and then expand. Big-bang launches hide problems until they are expensive.
Integrate or Redesign Workflows
You can augment an existing process by dropping Claude into one step, or redesign the process around what Claude makes possible. Augmenting is faster to adopt. Redesigning can unlock bigger gains but needs more change management.
Stakeholder Communication
When someone asks 'can Claude do this?', give an honest answer that covers value and limits. The best answer describes the division of labour: what Claude does, what people do, and where the review step sits.
Domain 5
12% of examConfiguration and Knowledge Management
Set up Projects so Claude starts every conversation informed: know which layer holds what, choose between uploads and live connectors, write short standing instructions, and keep configurations current.
Projects: Instructions and Knowledge
A Project has layers, and each layer has a job. Project instructions hold the standing rules that apply to every conversation. Project knowledge holds reference material. The individual prompt holds what is specific to this one request.
Uploaded Knowledge and Connectors
Information can reach Claude by upload or by connector. An uploaded file is a snapshot that goes stale. A connector, such as to a drive or mailbox, reads current content when you use it. Choose based on how often the source changes.
System-Level and Project Instructions
Standing instructions work best when they are short, specific, and maintained. A tight list of rules Claude can follow beats a long essay of preferences, and instructions need an owner who updates them as the work changes.
Maintain and Update Configurations
A Project is a living asset. You maintain its knowledge, and stale or cluttered configurations produce confidently wrong results. Replace outdated files, prune what no longer applies, and assign clear ownership for quality.
Domain 6
15% of examGovernance, Risk, and Responsible Use
Use Claude within the rules: separate what it can do from what you may do, protect sensitive data, follow organisational policy, and take responsibility for ethical outcomes.
Appropriate vs Inappropriate Use
Capability is not permission. Claude being able to do a task does not mean you are allowed to use it for that task, with that data, in that way. Know the limits of appropriate use and escalate when a need goes beyond the chat product.
Data Sensitivity, Privacy, and Regulation
Protect sensitive data before it reaches Claude. The reliable pattern is anonymise first, then analyse. Telling Claude not to retain or reuse information is not a control; removing the sensitive details beforehand is.
Organizational AI Policy
Your organisation's AI policy decides what is allowed. Read it before you paste anything, follow it even when it is inconvenient, and escalate when the policy is silent or ambiguous rather than guessing.
Ethical Implications
Using AI at work carries ethical duties. Be honest about when AI helped, keep a named person accountable for results, watch for bias, and keep humans in charge of decisions that materially affect people.
Domain 7
10% of examTroubleshooting and Optimization
When results disappoint, diagnose before rewriting: map symptom to cause to first fix, change one variable at a time, and make workflows cheaper and faster without losing quality.
Diagnose Underperforming Outputs
Before rewriting a prompt, work out what is actually wrong. Move from symptom to cause to first fix. Generic output, wrong facts, ignored instructions, and drift each point to a different cause and a different first remedy.
Adjust from Feedback
When you change several things at once, you cannot tell which change helped. Adjust one variable at a time, test, and keep what works. Use feedback from reviewers and users as the signal for what to change next.
Optimize Workflows
Once a workflow works, make it cheaper and faster without lowering quality. Share what works through Projects, standardise good prompts, and use the smallest model that still meets the bar.
Ready to practice?
Put the curriculum to work with hands-on exercises and a full mock exam.