Anthropic CCA-F (Claude Certified Architect - Foundations (CCAR-F)) Practice Test 23

Question 220

A release skill should be available through an explicit slash command but should not be selected automatically by the model. Which skill setting matches that intent?

  1. A skill name beginning with manual while leaving invocation settings unchanged.
  2. context: fork, with no change to invocation controls.
  3. user-invocable: false
  4. disable-model-invocation: true
  5. disable-model-invocation: false together with user-invocable: true.

Explanation

Correct Answer

D:

This setting keeps model-initiated invocation disabled while allowing the user to request the skill. Explicit invocation remains a way for the user to request the procedure. This selection setting does not replace any separate authorization checks required by the release actions.

Incorrect Answers

A:

A naming convention describes intent but does not disable model-initiated invocation.

B:

Forked execution controls the context in which a skill runs. It does not make the skill available only through a user's explicit command.

C:

Hiding user invocation is the opposite of preserving an explicit user command.

E:

These values leave both model and user invocation available. They do not impose the required restriction on automatic selection.

Question 221

A team-specific deployment checklist must be versioned with one repository. Where is the most suitable skill location?

  1. Only in one developer's personal skill directory.
  2. In an untracked temporary directory referenced only by the current shell session.
  3. Only in a global skill directory managed independently of this repository's revisions.
  4. In the repository's .claude/skills directory.

Explanation

Correct Answer

D:

A project skill can travel with the repository and be reviewed alongside the deployment workflow. The project location makes the checklist available in the intended repository rather than as a personal global procedure. Supporting files can be versioned with the skill for reproducible use.

Incorrect Answers

A:

A personal installation will not automatically provide the team's repository-specific procedure to colleagues.

B:

A temporary personal location will not travel with the repository's version history. Teammates cannot reliably obtain the checklist through a normal checkout.

C:

Global distribution can be useful for shared personal skills, but it does not bind this team's checklist to the requested repository version.

Question 222

A skill works for its author but fails in another checkout because it reads /home/alex/project/checklist.txt. What is the best portability fix?

  1. Require all checkouts to use the author's absolute directory as a deployment prerequisite.
  2. Paste a different absolute path into the skill for each teammate.
  3. Package the checklist with the skill and resolve it relative to the skill's location.
  4. Remove the checklist step without reviewing its purpose.

Explanation

Correct Answer

C:

Bundled relative resources remove dependence on the author's personal filesystem layout. References should resolve from a known skill location rather than the caller's incidental working directory. That lets the package move between checkouts without per-developer path edits.

Incorrect Answers

A:

A fixed personal path unnecessarily couples a reusable skill to one filesystem layout.

B:

Per-person copies create maintenance drift instead of a portable package.

D:

Deleting required procedure content hides the path problem rather than solving it.

Question 223

A release skill accepts a target environment as an argument. What should it do when the user supplies an unknown environment?

  1. Treat every unknown value as production.
  2. Use whichever environment appears first in a directory listing.
  3. Interpolate the raw argument into a shell command.
  4. Validate against supported environments and request or report a valid target before proceeding.

Explanation

Correct Answer

D:

Explicit argument validation prevents an ambiguous parameter from silently selecting an operational target. A deployment target is a semantic choice, not merely an arbitrary string to interpolate. Rejecting an unknown value avoids silently directing the release at a more consequential environment.

Incorrect Answers

A:

Defaulting an invalid target to production can direct the workflow at the wrong environment.

B:

Filesystem ordering does not represent the user's intended deployment destination.

C:

Unvalidated input can produce an unintended command or target selection.

Question 224

A Claude Code skill should insert all text following its slash command into a designated instruction slot. Which documented placeholder serves that purpose?

  1. The skill's description field alone.
  2. $ARGUMENTS
  3. $HOME
  4. The literal name of the most recent repository branch.

Explanation

Correct Answer

B:

This placeholder expands the invocation's argument text at the intended place in the skill content. Insertion conveys the user's supplied issue number or other arguments into the procedure. It does not remove the need to validate arguments before using them in tools or shell commands.

Incorrect Answers

A:

A description helps identify the skill but is not a placeholder for invocation text.

C:

This convention identifies a home directory, not the skill invocation's argument payload.

D:

A branch name does not represent the caller's supplied skill arguments.

Question 225

A shared review skill changes its definition of a blocking finding. How should a team roll out that change?

  1. Publish the change only in a terminal notification.
  2. Version the change and test representative borderline reviews against the revised criterion.
  3. Assume existing evaluation results remain valid without rerunning relevant cases.
  4. Silently overwrite all past review reports with the new wording.

Explanation

Correct Answer

B:

A recorded revision and targeted examples reveal how the new rule changes actual classifications. Borderline cases are especially useful because they expose where the new blocking threshold differs. Past reports should remain associated with the criteria actually used when they were produced.

Incorrect Answers

A:

A transient notification does not provide a reviewable definition or evidence of its effects.

C:

Changing the decision criterion can alter outcomes even if the surrounding workflow is unchanged.

D:

Historical reports should not be rewritten to imply they used a different decision rule.

Question 226

A team wants a formatter to run after Claude Code successfully edits a source file. Which hook timing fits this purpose?

  1. A matching PostToolUse hook for the editing operation.
  2. A hook that runs only when a user signs in.
  3. A PreToolUse hook that formats only the old file and never runs again.
  4. A final prompt asking the model to remember formatting next month.

Explanation

Correct Answer

A:

Post-action timing lets the formatter operate on the file content produced by the edit. The matcher should target the editing tools and file types relevant to the formatter. A failed edit should not be mistaken for a successfully updated file requiring postprocessing.

Incorrect Answers

B:

Authentication timing does not track source-file edits during the session.

C:

Formatting before the change cannot correct formatting introduced by the subsequent edit.

D:

A future reminder does not automate formatting after each relevant edit.

Question 227

A hook receives tool input as JSON, including a file path that may contain spaces. What is the reliable way to inspect that path?

  1. Search the entire input for the first slash and copy everything after it.
  2. Take the first string value from the JSON regardless of its property name.
  3. Split the raw JSON on whitespace and take the final token.
  4. Parse the JSON and read the documented path field for the matched tool.

Explanation

Correct Answer

D:

Structured parsing preserves the field value and avoids guessing from raw text formatting. JSON field access preserves spaces and escaped characters within the path. The handler should also account for differences in input shape between the tools it matches.

Incorrect Answers

A:

A slash may occur in unrelated data, and this method ignores field boundaries.

B:

The first string can be a session ID or command rather than the intended file path.

C:

Whitespace splitting breaks valid paths and does not respect JSON structure.

Question 228

A team adds a formatting hook but accidentally runs it after every tool call, including searches. What should it change first?

  1. Increase the formatter timeout on all events.
  2. Keep the broad matcher and return success whenever the formatter receives a search-result payload.
  3. Narrow the hook matcher and verify the relevant input fields before formatting.
  4. Run the formatter on the last remembered path whenever the current tool call supplies no file path.

Explanation

Correct Answer

C:

A targeted matcher limits the action to tool events that can create files needing formatting. A tool event can be relevant to search, reads, or mutations, so an unqualified hook runs too broadly. Matching the intended operation avoids unnecessary formatting work and incidental side effects.

Incorrect Answers

A:

More time does not make a formatter relevant to search-only operations.

B:

Ignoring unsuitable payloads can mask unintended hook execution. The matching rule should select the edit operations that actually require formatting.

D:

A stale path can cause unrelated files to be reformatted after searches. The hook should match the intended event and validate that event's input.

Question 229

An edit event can trigger a formatter more than once during retries. Which property is most valuable for the formatter hook?

  1. Deleting the file before determining whether it needs formatting.
  2. Producing the same formatted result when rerun on already formatted content.
  3. Appending a new copyright line on every invocation.
  4. Rewriting the file with a fresh timestamp comment after every successful format pass.

Explanation

Correct Answer

B:

Idempotent formatting tolerates repeated events without introducing additional edits. Once the file satisfies the formatting rules, another pass should leave it unchanged. This property makes retries and repeated event delivery harmless for the formatter's output.

Incorrect Answers

A:

Destructive preprocessing is not necessary to make repeated formatting reliable.

C:

Repeated invocation would accumulate duplicate changes instead of stabilizing the file.

D:

A changing timestamp creates new differences each time and prevents stable repeat execution.

Question 230

A required validation hook cannot find its checker executable. The team wants failures to be actionable. What should the hook provide?

  1. A clear diagnostic naming the missing dependency and a failure result consistent with the required gate.
  2. The application's production secrets to help debugging.
  3. An empty response that the team interprets as a passed check.
  4. A success result followed by an undocumented warning in a temporary file.

Explanation

Correct Answer

A:

The workflow needs both an enforceable failure signal and information that lets the team repair the environment. The failure signal should match the hook event's documented gating behavior. A success exit would conceal the fact that the required validation executable never ran.

Incorrect Answers

B:

Credential disclosure does not explain or repair a missing local executable.

C:

Silence cannot establish that the required checker completed successfully.

D:

Reporting success lets the required gate pass while hiding the reason validation never ran.

Claude Certified Architect - Foundations (CCAR-F) (CCA-F) Practice Tests