The Four Project Configuration Mechanisms for the CCAO-F Exam

From using Claude to operating a Project

There is a line between using Claude and operating it. Using it means typing one good prompt today. Operating it means building an environment so that every new conversation starts from a configured baseline rather than a blank slate. In claude.ai that environment is a Project, and the Claude Certified Associate - Foundations (CCAO-F) exam treats knowing its parts as a foundational, remember-level skill.

The reason this matters is that configuration is leverage. You set it up once and benefit on every conversation that follows, and two teammates asking the same question get the same quality of answer instead of each re-inventing context daily. But that only works if you understand that a Project is not a single settings box. It is four separate slots, and each one is built for a different kind of recurring need.

What each of the four slots is for

Standing instructions govern behavior. They express how Claude should act across every conversation in the Project: the tone it uses, the formats it defaults to, the verification habits it follows. They are rules about conduct, not statements of fact.

The knowledge base holds facts. It is where the documents, policies, and reference files that Claude should draw on live, so nobody has to re-upload a brand guide or a pricing table each time. It answers "what is true," not "how to behave."

Skills carry procedures. A Skill encodes a repeatable, multi-step task that Claude should perform consistently, such as turning findings into a standard report layout. Crucially, Skills are built once at the account level and reused across any Project that needs them, so a procedure is not duplicated Project by Project.

Scoped Memory keeps continuity. It preserves what a Project has already settled: the stakeholder names, the standing preferences, the decisions from earlier conversations you would rather not restate. The word that matters is scoped: a Project's Memory is sectioned off from your other Projects.

Why "configuration" is four things, not one

The most common early mistake is to treat configuration as a single dial you turn up or down. It is not. A behavior rule, a reference fact, a procedure, and a piece of continuity are four different kinds of thing, and each has a slot suited to it. The instruction-versus-fact distinction between the first two slots is the one candidates get wrong most often, but the four-way split is the frame the whole domain rests on.

Slots inside a Project, layers across the platform

It helps to hold two framings at once. Inside a single Project, the four mechanisms are configuration slots you fill in. Stepping back, the platform capabilities they draw on behave as independent, composable layers: the Project itself carries standing instructions and a knowledge base, Skills add reusable procedures, Code Execution lets Claude run code in a sandbox to compute or transform data, and Memory preserves scoped continuity. The point is that these layers are not welded together. Each can be switched on or off on its own, and they stack rather than compete: a Project can lean on Skills without Code Execution, or use Memory and knowledge without any Skill at all. Thinking in layers keeps you from assuming that enabling one capability drags the others along, and it explains why routing a need to the wrong slot rarely fails loudly. The layers still run; they just run on the wrong instructions. Code Execution in particular is a distinct capability rather than something baked into instructions or knowledge, and it composes with the other three when a task needs real computation instead of prose.

What the CCAO-F exam trips candidates on

The exam probes two confusions here. The first is treating configuration as one setting rather than four mechanisms, each suited to a different need. A question will describe several distinct needs at once and reward the answer that maps each to its own slot rather than lumping them together.

The second is assuming the knowledge base and Memory are interchangeable because both "store information." They are not. The knowledge base holds stable reference material that you want to keep current and consistent; Memory holds an evolving record of what the Project has decided. Confusing the two makes a Project harder to maintain, because a stable fact parked in Memory is harder to keep authoritative, and an evolving decision buried in a document is easy to miss.

Common misreadings to avoid

How this shows up on the exam

Domain 5 questions on this knowledge point list a set of recurring needs and ask which mechanism handles each. The reliable move is to name the four slots and match by type: behavior to instructions, facts to knowledge, procedures to Skills, continuity to Memory.

This foundation feeds directly into instructions govern behavior while knowledge holds facts, which sharpens the first two slots, and into Skills as account-level reusable procedures and scoped Memory and Project isolation. Once the four slots are clear, the harder analysis questions about selecting the right mechanism become tractable.