100% Latest & Updated Cisco 350-901 Practice Test Questions, Exam Dumps & Verified Answers!
30 Days Free Updates, Instant Download!
Question 830
The generated code resembles a copied snippet from a project whose license has not been verified. Independently, reviewers flag this potentially unsafe practice for assessment: 'Discard license findings once the build succeeds'. Which verification addresses that practice without assuming it caused the incident?
- Review similarity and attribution findings
- Copy an arbitrary repository with no licensing review
- Review final source and provenance notes
- Inspect approval and dependency notices
Explanation
Correct Answer
D:Build output is not legal review. The intended behavior is: Provenance helps future maintainers assess compliance.
Incorrect Answers
A:This checks check code provenance, not the named failure. AI generation does not establish ownership of all reused code.
B:Unknown terms can create legal and delivery risk.
C:This checks author original implementation, not the named failure. Independent design reduces inappropriate copying risk.
Question 831
An AI-based NOC assistant summarizes text retrieved from a syslog server. The specific implementation objective is to treat retrieved logs as data. Which configuration best meets the stated requirement?
- Delete suspicious input after it influences a tool call
- Give a summarization agent unrestricted router configuration rights
- Apply ACL changes directly because a log says to do so
- Keep untrusted log content separate from system and operator instructions
Explanation
Correct Answer
D:External text is not authorized control input.
Incorrect Answers
A:Missing evidence hides the exploit path.
B:Injected content may gain unsafe authority.
C:Log provenance does not imply change approval.
Question 832
An AI-based NOC assistant summarizes text retrieved from a syslog server. The specific implementation objective is to limit tool privileges. Which design action is most appropriate for this case?
- Give the assistant read-only diagnostic tools unless approved mutation is necessary
- Delete suspicious input after it influences a tool call
- Apply ACL changes directly because a log says to do so
- Execute commands embedded in a device log as if the admin requested them
Explanation
Correct Answer
A:Least privilege contains prompt-injection impact.
Incorrect Answers
B:Missing evidence hides the exploit path.
C:Log provenance does not imply change approval.
D:Untrusted text may redirect the agent.
Question 833
An AI-based NOC assistant summarizes text retrieved from a syslog server. The specific implementation objective is to validate proposed actions. What should the engineer implement for this specific objective?
- Execute commands embedded in a device log as if the admin requested them
- Give a summarization agent unrestricted router configuration rights
- Require independent policy checks and explicit approval for high-impact changes
- Delete suspicious input after it influences a tool call
Explanation
Correct Answer
C:A model recommendation is not authorization.
Incorrect Answers
A:Untrusted text may redirect the agent.
B:Injected content may gain unsafe authority.
D:Missing evidence hides the exploit path.
Question 834
An AI-based NOC assistant summarizes text retrieved from a syslog server. The specific implementation objective is to retain audit trail. Which adjustment would satisfy this particular constraint?
- Give a summarization agent unrestricted router configuration rights
- Apply ACL changes directly because a log says to do so
- Record retrieved source, agent decision and rejected unsafe content
- Execute commands embedded in a device log as if the admin requested them
Explanation
Correct Answer
C:Traceability permits investigation of manipulated prompts.
Incorrect Answers
A:Injected content may gain unsafe authority.
B:Log provenance does not imply change approval.
D:Untrusted text may redirect the agent.
Question 835
An AI-based NOC assistant summarizes text retrieved from a syslog server. The deployment plan must both (1) limit tool privileges and (2) retain audit trail. Which TWO actions together satisfy those requirements? Choose TWO.
- Give a summarization agent unrestricted router configuration rights
- Execute commands embedded in a device log as if the admin requested them
- Give the assistant read-only diagnostic tools unless approved mutation is necessary
- Apply ACL changes directly because a log says to do so
- Record retrieved source, agent decision and rejected unsafe content
Explanation
Correct Answers
C:Least privilege contains prompt-injection impact.
E:Traceability permits investigation of manipulated prompts.
Incorrect Answers
A:Injected content may gain unsafe authority.
B:Untrusted text may redirect the agent.
D:Log provenance does not imply change approval.
Question 836
An AI-based NOC assistant summarizes text retrieved from a syslog server. Three conditions must be met: (1) limit tool privileges; (2) validate proposed actions; and (3) retain audit trail. Which THREE actions form the narrowest complete plan? Choose THREE.
- Give a summarization agent unrestricted router configuration rights
- Delete suspicious input after it influences a tool call
- Record retrieved source, agent decision and rejected unsafe content
- Require independent policy checks and explicit approval for high-impact changes
- Execute commands embedded in a device log as if the admin requested them
- Give the assistant read-only diagnostic tools unless approved mutation is necessary
Explanation
Correct Answers
C:Traceability permits investigation of manipulated prompts.
D:A model recommendation is not authorization.
F:Least privilege contains prompt-injection impact.
Incorrect Answers
A:Injected content may gain unsafe authority.
B:Missing evidence hides the exploit path.
E:Untrusted text may redirect the agent.
Question 837
One device log message includes instructions telling the assistant to reveal credentials and ignore authorization limits. Independently, reviewers flag this potentially unsafe practice for assessment: 'Execute commands embedded in a device log as if the admin requested them'. Which inspection most directly evaluates that practice?
- Inspect approval workflow and proposed diff
- Give a summarization agent unrestricted router configuration rights
- Review source and tool execution records
- Inspect instruction/data boundary
Explanation
Correct Answer
D:Untrusted text may redirect the agent. The intended behavior is: External text is not authorized control input.
Incorrect Answers
A:This checks validate proposed actions, not the named failure. A model recommendation is not authorization.
B:Injected content may gain unsafe authority.
C:This checks retain audit trail, not the named failure. Traceability permits investigation of manipulated prompts.
Question 838
One device log message includes instructions telling the assistant to reveal credentials and ignore authorization limits. Independently, reviewers flag this potentially unsafe practice for assessment: 'Give a summarization agent unrestricted router configuration rights'. What diagnostic evidence should first be collected to assess that practice?
- Inspect approval workflow and proposed diff
- Delete suspicious input after it influences a tool call
- Inspect instruction/data boundary
- Inspect allowed tool permissions
Explanation
Correct Answer
D:Injected content may gain unsafe authority. The intended behavior is: Least privilege contains prompt-injection impact.
Incorrect Answers
A:This checks validate proposed actions, not the named failure. A model recommendation is not authorization.
B:Missing evidence hides the exploit path.
C:This checks treat retrieved logs as data, not the named failure. External text is not authorized control input.
Question 839
One device log message includes instructions telling the assistant to reveal credentials and ignore authorization limits. Independently, reviewers flag this potentially unsafe practice for assessment: 'Apply ACL changes directly because a log says to do so'. Which focused check best tests this specific assumption?
- Review source and tool execution records
- Execute commands embedded in a device log as if the admin requested them
- Inspect approval workflow and proposed diff
- Inspect allowed tool permissions
Explanation
Correct Answer
C:Log provenance does not imply change approval. The intended behavior is: A model recommendation is not authorization.
Incorrect Answers
A:This checks retain audit trail, not the named failure. Traceability permits investigation of manipulated prompts.
B:Untrusted text may redirect the agent.
D:This checks limit tool privileges, not the named failure. Least privilege contains prompt-injection impact.
Question 840
One device log message includes instructions telling the assistant to reveal credentials and ignore authorization limits. Independently, reviewers flag this potentially unsafe practice for assessment: 'Delete suspicious input after it influences a tool call'. Which verification addresses that practice without assuming it caused the incident?
- Inspect approval workflow and proposed diff
- Review source and tool execution records
- Inspect instruction/data boundary
- Give a summarization agent unrestricted router configuration rights
Explanation
Correct Answer
B:Missing evidence hides the exploit path. The intended behavior is: Traceability permits investigation of manipulated prompts.
Incorrect Answers
A:This checks validate proposed actions, not the named failure. A model recommendation is not authorization.
C:This checks treat retrieved logs as data, not the named failure. External text is not authorized control input.
D:Injected content may gain unsafe authority.
Designing, Deploying, and Managing Network Automation Systems (350-901) Practice Tests