Question 190
A staging environment silently points at production data through a shared MCP URL. What configuration check would prevent this?
- Store both version identifiers and compare them on the same evaluation set before switching traffic.
- Validate environment-specific endpoint mappings and credentials before deployment.
- Test connector compatibility without checking the staging environment endpoint or credentials.
- Deploy schema and consumer changes with compatibility tests under a coordinated release.
- Choose environment-appropriate settings and restrict the CI permission mode.
Explanation
Correct Answer
B:The correct tool schema can still cause damage when its endpoint is misconfigured; environment segregation must include targets and identity.
Incorrect Answers
A:Behavior changes across model versions are deployment changes; a tested rollout prevents silent production regressions; it changes multi-criterion integration; it does not prove environment endpoint isolation in this case.
C:Dependency compatibility does not establish which environment the application connects to. A connector can pass every schema test while its staging endpoint still points at production.
D:Generated output contracts span both sides of the interface; configuration drift can break consumers even when API calls succeed; the consequence is improved schema release coordination, not verified environment endpoint isolation.
E:Automation runs without the same human supervision as local sessions, so its authority should be explicitly narrower; where environment endpoint isolation matters, a control for environment-specific Claude Code permissions leaves that test unmet.
Question 191
A release engineer changes a prompt and the model version used by a Claude application. Which TWO records should be managed together to make failures reproducible? Choose TWO.
- Record the selected model identifier and controlled rollout status.
- Persist the exact prompt revision used for each deployment.
- Disable integration tests during model migration.
- Keep only the latest natural-language user response.
- Assume the repository commit alone captures the active model.
Explanation
Correct Answers
A:Model behavior and supported parameters may change across versions; explicit identity enables comparison and rollback. An explicit model version prevents later investigations from confusing alias changes with prompt regressions.
B:Prompt text shapes model behavior and must be traceable to an incident or regression test. Include template parameters and the effective deployment environment, since those can alter the rendered prompt.
Incorrect Answers
C:Removing tests reduces change assurance and does not improve reproducibility. Migration tests are needed to isolate incompatibilities in tool use, prompts, and structured-output handling.
D:A final reply does not identify which model and prompt produced it. Final prose alone cannot reconstruct the system instructions and model configuration that produced the outcome.
E:External aliases and runtime configuration can differ from source code unless recorded. A deployed model alias or environment override may not match the version assumed from a code revision.
Question 192
A developer opens a new repository and wants a starting CLAUDE.md describing its structure. What interactive action is designed for this?
- Specify a Claude Code subagent with role instructions and permitted capabilities.
- Employ /model and select a supported model.
- Set up deny rules and suitable hook or sandbox policies in settings.
- Run /init in Claude Code and review the generated project guidance.
Explanation
Correct Answer
D:The initialization command creates a starter instruction file; the team should still edit it to reflect real repository conventions.
Incorrect Answers
A:Subagent definitions create isolated task contexts and can restrict tool access, separating review from broad editing authority; for initialize repo instructions, a control centered on configured code review subagent is not the appropriate substitute.
B:Model selection is a session configuration operation; rewriting a user prompt does not itself guarantee a runtime model change; even if switch active coding model succeeds, initialize repo instructions still needs separate verification.
C:Operational permission controls govern actual tool use more reliably than a statement inside CLAUDE.md alone; the consequence is improved deny risky file tools, not verified initialize repo instructions.
Question 193
A long Claude Code session has grown noisy, but the task should continue with a condensed history. Which command fits?
- Use /compact with guidance about facts to preserve.
- Use the supported resume workflow to reopen the saved conversation.
- Build a reusable Skill with documented instructions and supporting assets.
- Launch /mcp to view and manage server connections.
Explanation
Correct Answer
A:Compaction summarizes prior context for continued work, unlike starting an empty conversation that discards the working thread.
Incorrect Answers
B:Session continuation retrieves recorded history or state; merely recreating the same working directory does not recover a conversation; the scenario's controlling requirement is compact active session, not the related multi-criterion integration concern.
C:Skills package repeatable procedures and context for assistant use; they are not a replacement for executable tool permissions; a remedy for reusable Claude Code skill is not a substitute for compact active session.
D:The MCP command exposes connection configuration and authentication rather than editing the conversation's natural-language instructions; it follows inspect MCP connections constraints but leaves compact active session unspecified.
Question 194
A developer wants to start a separate conversation without retaining the current turn history in context. What command is appropriate?
- As an alternative, do not treat each stream event as a complete assistant answer.
- Employ /compact with guidance about facts to preserve.
- Establish a Claude Code subagent with role instructions and permitted capabilities.
- Use /clear to begin a fresh session context.
Explanation
Correct Answer
D:Clearing starts an empty conversation; it is different from a summary-based compacting operation for continued tasks.
Incorrect Answers
A:Incremental event streams contain intermediate messages and metadata; the consumer must recognize completion before using final results; a remedy for interpret headless stream events is not a substitute for clear conversational context.
B:Compaction summarizes prior context for continued work, unlike starting an empty conversation that discards the working thread; the clear conversational context criterion differs from compact active session; the gap persists.
C:Subagent definitions create isolated task contexts and can restrict tool access, separating review from broad editing authority; for clear conversational context, this instead manages configured code review subagent; the requested decision remains unresolved.
Question 195
A CI job must invoke Claude Code without an interactive terminal and consume a machine-readable result. What interface should it use?
- Structure scoped rules so relevant directories receive the appropriate guidance.
- Employ /clear to begin a fresh session context.
- Run Claude Code in print/headless mode with a supported structured output format.
- Adjust deny rules and suitable hook or sandbox policies in settings.
- Document a Claude Code subagent with role instructions and permitted capabilities.
Explanation
Correct Answer
C:Noninteractive invocation is suited to automation; the caller must still parse completion status and restrict tool permissions.
Incorrect Answers
A:Path-specific rules reduce instruction clutter and make different coding standards applicable in the appropriate project area; this targets scoped repository rules, while the system must demonstrate headless CI invocation.
B:Clearing starts an empty conversation; it is different from a summary-based compacting operation for continued tasks; where headless CI invocation matters, a control for clear conversational context leaves that test unmet.
D:Operational permission controls govern actual tool use more reliably than a statement inside CLAUDE.md alone; even if deny risky file tools succeeds, headless CI invocation still needs separate verification.
E:Subagent definitions create isolated task contexts and can restrict tool access, separating review from broad editing authority; it improves configured code review subagent only, leaving headless CI invocation unproven.
Question 196
A team wants standard instructions for repeated dependency upgrades that the assistant can invoke when relevant. Which Claude Code customization fits?
- Execute /init in Claude Code and review the generated project guidance.
- Create a reusable Skill with documented instructions and supporting assets.
- Use /mcp to view and manage server connections.
- Employ the supported resume workflow to reopen the saved conversation.
Explanation
Correct Answer
B:Skills package repeatable procedures and context for assistant use; they are not a replacement for executable tool permissions.
Incorrect Answers
A:The initialization command creates a starter instruction file; the team should still edit it to reflect real repository conventions; for reusable Claude Code skill, this instead manages initialize repo instructions; the requested decision remains unresolved.
C:The MCP command exposes connection configuration and authentication rather than editing the conversation's natural-language instructions; this may help inspect MCP connections but cannot by itself deliver reusable Claude Code skill.
D:Session continuation retrieves recorded history or state; merely recreating the same working directory does not recover a conversation; the effect concerns multi-criterion integration; the success condition is reusable Claude Code skill.
Question 197
A repository needs distinct coding conventions for infrastructure files versus frontend components. What instruction structure helps?
- Set deny rules and suitable hook or sandbox policies in settings.
- Adopt /compact with guidance about facts to preserve.
- Organize scoped rules so relevant directories receive the appropriate guidance.
- Perform /init in Claude Code and review the generated project guidance.
Explanation
Correct Answer
C:Path-specific rules reduce instruction clutter and make different coding standards applicable in the appropriate project area.
Incorrect Answers
A:Operational permission controls govern actual tool use more reliably than a statement inside CLAUDE.md alone; testing scoped repository rules would still fail after improving deny risky file tools.
B:Compaction summarizes prior context for continued work, unlike starting an empty conversation that discards the working thread; for scoped repository rules, a control centered on compact active session is not the appropriate substitute.
D:The initialization command creates a starter instruction file; the team should still edit it to reflect real repository conventions; it improves initialize repo instructions only, leaving scoped repository rules unproven.
Question 198
A developer wants a specialized code-review worker with its own prompt and restricted tools. What should be configured?
- Execute Claude Code in print/headless mode with a supported structured output format.
- Carry out /init in Claude Code and review the generated project guidance.
- Define a Claude Code subagent with role instructions and permitted capabilities.
- Develop a reusable Skill with documented instructions and supporting assets.
Explanation
Correct Answer
C:Subagent definitions create isolated task contexts and can restrict tool access, separating review from broad editing authority.
Incorrect Answers
A:Noninteractive invocation is suited to automation; the caller must still parse completion status and restrict tool permissions; where configured code review subagent matters, a control for headless CI invocation leaves that test unmet.
B:The initialization command creates a starter instruction file; the team should still edit it to reflect real repository conventions; changing initialize repo instructions does not establish configured code review subagent in this application.
D:Skills package repeatable procedures and context for assistant use; they are not a replacement for executable tool permissions; the effect concerns reusable Claude Code skill; the success condition is configured code review subagent.
Question 199
An engineer needs to inspect which MCP servers are connected in a Claude Code session. Which command is directly relevant?
- For this workflow, do not treat each stream event as a complete assistant answer.
- Open /mcp to view and manage server connections.
- Perform Claude Code in print/headless mode with a supported structured output format.
- Introduce a reusable Skill with documented instructions and supporting assets.
Explanation
Correct Answer
B:The MCP command exposes connection configuration and authentication rather than editing the conversation's natural-language instructions.
Incorrect Answers
A:Incremental event streams contain intermediate messages and metadata; the consumer must recognize completion before using final results; the effect concerns interpret headless stream events; the success condition is inspect MCP connections.
C:Noninteractive invocation is suited to automation; the caller must still parse completion status and restrict tool permissions; testing inspect MCP connections would still fail after improving headless CI invocation.
D:Skills package repeatable procedures and context for assistant use; they are not a replacement for executable tool permissions; because reusable Claude Code skill is its concern, inspect MCP connections remains outside its effect here.
Question 200
A teammate wants to change the model for an ongoing Claude Code session. Which built-in control is intended for this?
- Access /mcp to view and manage server connections.
- Set out a Claude Code subagent with role instructions and permitted capabilities.
- Use /model and select a supported model.
- Arrange scoped rules so relevant directories receive the appropriate guidance.
Explanation
Correct Answer
C:Model selection is a session configuration operation; rewriting a user prompt does not itself guarantee a runtime model change.
Incorrect Answers
A:The MCP command exposes connection configuration and authentication rather than editing the conversation's natural-language instructions; this implements inspect MCP connections rather than the demanded switch active coding model.
B:Subagent definitions create isolated task contexts and can restrict tool access, separating review from broad editing authority; when the goal is switch active coding model, optimizing configured code review subagent misses that operational boundary.
D:Path-specific rules reduce instruction clutter and make different coding standards applicable in the appropriate project area; the consequence is improved scoped repository rules, not verified switch active coding model.