Anthropic CCDV-F (Claude Certified Developer - Foundations) Practice Test 39

Question 380

An MCP server publishes a reusable prompt named "draft-incident-summary" that expects an incident_id argument. How should a client request the filled-out template?

  1. Invoke a shell command named after the prompt without querying its advertised arguments.
  2. Ask the language model to guess the required incident identifier without consulting the caller.
  3. Call resources/read with an arbitrary incident_id and assume it executes the prompt template.
  4. Request the named prompt with its defined arguments and consume the returned prompt messages.

Explanation

Correct Answer

D:

MCP prompt templates can expose arguments that the client supplies when requesting a prompt. The server generates the corresponding prompt content; this is distinct from invoking an executable tool to mutate an incident.

Incorrect Answers

A:

A prompt is an instruction template, not automatically a shell tool. Treating it as an executable command introduces an unauthorized capability that the template did not define.

B:

A required prompt argument must correspond to the intended record. Guessing the identifier risks using the wrong incident and undermines reliable template parameterization.

C:

Resource reads retrieve identified content rather than expand a named prompt with arguments. The client should use the prompt retrieval contract for a template.

Question 381

A new remote MCP server authenticates all callers with one publicly shared token. What should deployment change?

  1. Employ an appropriate authenticated HTTP-based MCP transport.
  2. Handle returned notes as untrusted content and enforce tool permissions independently.
  3. Publish the capability through a reusable MCP server with a documented tool contract.
  4. Require per-client or appropriately scoped authenticated access at the server boundary.

Explanation

Correct Answer

D:

MCP protocol compliance is not authorization; network-exposed capabilities need identity, transport security, and permission enforcement.

Incorrect Answers

A:

A remote server needs network reachability and access controls; stdio is designed for locally connected processes; the scenario's controlling requirement is MCP remote server identity, not the related remote MCP transport concern.

B:

Protocol delivery does not elevate the trust of source data; embedded text can still contain prompt-injection instructions; even if trust boundary for MCP result succeeds, MCP remote server identity still needs separate verification.

C:

MCP provides a client-server integration boundary so multiple clients can discover and invoke the same service; where MCP remote server identity matters, a control for reusable cross-client tool leaves that test unmet.

Question 382

An MCP server silently changes the required input field for a published tool. Some connected clients still send the old field and begin failing. Which maintenance practice would have reduced the breakage?

  1. Replace the tool with an MCP prompt template but leave the executable endpoint unchanged.
  2. Keep the new schema undocumented so models learn it from trial and error.
  3. Ignore failures because MCP transport successfully delivered the requests.
  4. Version and test the tool contract, preserve compatibility when possible, and notify clients of discovery changes.

Explanation

Correct Answer

D:

MCP tool schemas are client-facing contracts. Breaking argument changes should be managed with migration tests and appropriate tool-list change signaling or rediscovery support for compatible clients.

Incorrect Answers

A:

Prompts provide reusable interaction templates, not a compatibility layer for a callable operation. Existing clients would still send invalid arguments to the changed tool.

B:

Guessing parameter names creates avoidable failed calls. Clients rely on tool metadata and must be able to refresh a documented, compatible interface.

C:

Successful transport does not validate tool input compatibility or business outcomes. The server must treat schema changes as an API lifecycle concern.

Question 383

A remote MCP server returns trusted business data and occasionally contains user-written notes. What protection does the consuming application still need?

  1. Adopt an appropriate authenticated HTTP-based MCP transport.
  2. Insist on per-client or appropriately scoped authenticated access at the server boundary.
  3. Treat returned notes as untrusted content and enforce tool permissions independently.
  4. Offer reference data as resources and executable capabilities as tools.

Explanation

Correct Answer

C:

Protocol delivery does not elevate the trust of source data; embedded text can still contain prompt-injection instructions.

Incorrect Answers

A:

A remote server needs network reachability and access controls; stdio is designed for locally connected processes; the consequence is improved remote MCP transport, not verified trust boundary for MCP result.

B:

MCP protocol compliance is not authorization; network-exposed capabilities need identity, transport security, and permission enforcement; for trust boundary for MCP result, this instead manages MCP remote server identity; the requested decision remains unresolved.

D:

Resources provide context to clients; tools are invoked to perform operations and may have side effects; when the goal is trust boundary for MCP result, optimizing MCP resources versus tools misses that operational boundary.

Question 384

A developer wants Claude to invoke an existing internal billing API from one application only. What customization is simplest?

  1. Apply a prompt or Skill to organize the tool's returned data.
  2. Set up a subagent rather than adding a generic web-search MCP tool.
  3. Define a narrow client-side custom tool for the application.
  4. Prefer the built-in when it satisfies functionality without extra trust boundaries.

Explanation

Correct Answer

C:

A direct tool definition fits an app-specific operation without the deployment complexity of a reusable server protocol.

Incorrect Answers

A:

Tool interfaces retrieve facts; reusable instructions define how those facts should be presented; for this request, one-app custom tool is the missing piece, not multi-criterion integration.

B:

A subagent creates an isolated reasoning role, whereas an MCP tool supplies an external capability; testing one-app custom tool would still fail after improving specialist agent versus external tool.

D:

New server deployments introduce authentication, latency, and maintenance overhead that may not add value; a remedy for avoid redundant custom integration is not a substitute for one-app custom tool.

Question 385

Several assistants use the same organization-wide MCP issue-search connector. One department must not read another department’s tickets. Where should cross-client access be enforced?

  1. Enforce the authenticated caller’s scope and object permissions inside the MCP-backed service.
  2. Ask users to promise not to search other departments’ tickets.
  3. Place each department’s name in the tool description and trust the model to stay within it.
  4. Give all assistants the same unrestricted connector token to avoid permission errors.

Explanation

Correct Answer

A:

Sharing an MCP connector does not imply sharing the same data authority. Server-side claims and record-level authorization must filter each tool call before results are exposed to any assistant.

Incorrect Answers

B:

Voluntary instructions cannot replace authentication and authorization. The backend must verify each requested object even when the user or model claims access is allowed.

C:

Descriptions help tool selection but are not an access-control decision. A caller could still request a restricted issue identifier through the same service endpoint.

D:

A broad shared credential removes the tenant boundary and lets one compromised integration access unrelated data. Isolation requires enforceable per-caller or per-scope permissions.

Question 386

A repeated code-review procedure needs instructions and reference documents but no new external executable function. Which mechanism fits?

  1. Package the procedure as an Agent Skill.
  2. Combine supported built-in and custom tools for actions with instructions for orchestration.
  3. Employ a tool for fetching and a reusable prompt or Skill for rubric application.
  4. Use again the service after checking its permissions, authentication, and tool contracts.

Explanation

Correct Answer

A:

Skills supply reusable instructions and assets, whereas tools represent callable capabilities with executable behavior.

Incorrect Answers

B:

No single Skill or MCP primitive replaces a complete agent design; each component should match the capability it provides; the consequence is improved mixed capability composition, not verified procedural customization.

C:

Retrieval is an executable integration, while evaluation criteria are instructions; separating them improves testability; this targets capability versus procedure separation, while the system must demonstrate procedural customization.

D:

Existing MCP connectivity saves implementation work only when its security and capabilities fit the application; even if evaluating existing MCP integration succeeds, procedural customization still needs separate verification.

Question 387

An agent must read and edit local files through supported platform capabilities. What should be checked before writing a new connector?

  1. Introduce the operation behind a secure MCP server or equivalent controlled service.
  2. Employ agents for choosing actions, tools for executing capabilities, and Skills for reusable procedures.
  3. Use an appropriate built-in tool when its features and permissions meet the task.
  4. Store the instructions and examples in a Skill or repository guidance.

Explanation

Correct Answer

C:

Built-ins reduce custom integration burden, but their scope must still be controlled. Compare available built-in permissions and platform limits before accepting the maintenance burden of custom connector code.

Incorrect Answers

A:

A server boundary can own credentials and permissions for all clients instead of copying privileged code into each prompt; for built-in file tool fit, this instead manages centralized tool service control; the requested decision remains unresolved.

B:

Separating planning, execution, and instruction assets reduces confused responsibility and unnecessary complexity; the scenario's controlling requirement is built-in file tool fit, not the related autonomy capability instruction distinction concern.

D:

Stable instructions can be invoked consistently without implementing a server-side action endpoint; it follows repeatable editing instructions constraints but leaves built-in file tool fit unspecified.

Question 388

A team publishes a Skill for migration reviews, but assistants rarely choose it when the user asks for that work. What part of Skill packaging should the authors improve?

  1. Deploy an unrelated MCP database server solely to make the Skill description longer.
  2. Rename the Skill to a random identifier without describing the workflow.
  3. Store credentials in the Skill description to prove it can reach every repository.
  4. Clarify the Skill’s description and triggering purpose so the assistant can recognize applicable migration tasks.

Explanation

Correct Answer

D:

A discoverable Skill needs metadata that communicates when its procedure is relevant. Clear boundaries and task vocabulary help the runtime select its instructions without turning it into an external execution service.

Incorrect Answers

A:

An external service is unnecessary for improving procedural discoverability. It adds a new capability boundary without fixing the instructions that identify when the Skill should apply.

B:

An opaque label gives little evidence of the task the package supports. Discovery depends on meaningful descriptions and documented purpose, not arbitrary uniqueness.

C:

Embedding secrets in discoverable instructional content broadens exposure and does not improve matching to migration-review requests. Authentication belongs in approved tool integrations.

Question 389

A business action needs an isolated, audited backend with explicit authentication and multiple clients. Which option best centralizes that boundary?

  1. Implement the operation behind a secure MCP server or equivalent controlled service.
  2. Specify a narrow client-side custom tool for the application.
  3. Bundle the procedure as an Agent Skill.
  4. Adopt agents for choosing actions, tools for executing capabilities, and Skills for reusable procedures.

Explanation

Correct Answer

A:

A server boundary can own credentials and permissions for all clients instead of copying privileged code into each prompt.

Incorrect Answers

B:

A direct tool definition fits an app-specific operation without the deployment complexity of a reusable server protocol; it changes one-app custom tool; it does not prove centralized tool service control in this case.

C:

Skills supply reusable instructions and assets, whereas tools represent callable capabilities with executable behavior; because procedural customization is its concern, centralized tool service control remains outside its effect here.

D:

Separating planning, execution, and instruction assets reduces confused responsibility and unnecessary complexity; the consequence is improved autonomy capability instruction distinction, not verified centralized tool service control.

Question 390

A developer wants Claude Code to use a specialist reviewer with its own instructions and tool restrictions. What customization should be chosen?

  1. Limit actual tool permissions rather than relying only on Skill text.
  2. Configure a subagent rather than adding a generic web-search MCP tool.
  3. Assemble supported built-in and custom tools for actions with instructions for orchestration.
  4. Employ an appropriate built-in tool when its features and permissions meet the task.

Explanation

Correct Answer

B:

A subagent creates an isolated reasoning role, whereas an MCP tool supplies an external capability. Give the specialist its own role-specific prompt and available tools; a generic external search endpoint does not create that role.

Incorrect Answers

A:

Tool availability and authorization define what can execute; a procedural document cannot enforce a security boundary; changing executable permission enforcement does not establish specialist agent versus external tool in this application.

C:

No single Skill or MCP primitive replaces a complete agent design; each component should match the capability it provides; where specialist agent versus external tool matters, a control for mixed capability composition leaves that test unmet.

D:

Built-ins reduce custom integration burden, but their scope must still be controlled; the consequence is improved built-in file tool fit, not verified specialist agent versus external tool.

Claude Certified Developer - Foundations (CCDV-F) Practice Tests