McAfee-Secured Website

Exam Bundle

Exam Code: CAS-005

Exam Name CompTIA SecurityX

Certification Provider: CompTIA

Corresponding Certification: CASP

CompTIA CAS-005 Bundle $44.99

100% CompTIA CAS-005 Guaranteed Success With Testking Training Materials

Get Real CAS-005 Exam Questions & Expert Verified Answers as Experienced on the Real Exam!

  • Questions & Answers

    CAS-005 Questions & Answers

    482 Questions & Answers

    The ultimate exam preparation tool, Questions & Answers mirrors the actual CAS-005 exam while allowing you to hone and measure every aspect of your test-taking ability.

  • CAS-005 Video Course

    CAS-005 Video Course

    180 Video Lectures

    CAS-005 Video Course is developed by CompTIA Professionals to help you pass the CAS-005 exam.

    Description

    This course will improve your knowledge and skills required to pass CompTIA SecurityX exam.
  • Study Guide

    CAS-005 Study Guide

    903 PDF Pages

    Developed by industry experts, this 903-page guide spells out in painstaking detail all of the information you need to ace CAS-005 exam.

CompTIA SecurityX (CAS-005) Practice Tests

Satisfaction Guaranteed

Money Back Guarantee

Testking provides hassle-free money back guarantee with our products. That is because we have 100% trust in the abilities of our professional and experience product team, and our record is a proof of that.

99.6% PASS RATE
Total Cost: $194.97
Bundle Price: $149.98

Purchase Individually

  • Questions & Answers

    Questions & Answers

    482 Questions

    $124.99
  • CAS-005 Video Course

    Video Course

    180 Video Lectures

    $39.99
  • Study Guide

    Study Guide

    903 PDF Pages

    $29.99

CompTIA CAS-005 Practice Test 52

Question 510

After the team responds to the following issue, which finding best demonstrates the desired security outcome? A service retrieves user-provided URLs on its backend and can contact cloud instance metadata addresses.

  1. The report separately records verified archive activity and any proven external transmission.
  2. Requests targeting metadata or private IP ranges fail while sanctioned external URLs still work.
  3. An altered or untrusted tooling package fails the signed provenance gate.
  4. A controlled reflected-XSS payload is displayed harmlessly without executing in the browser.

Explanation

Correct Answer

B:

This is direct test evidence of the control objective in the case: A service retrieves user-provided URLs on its backend and can contact cloud instance metadata addresses. It demonstrates that the proposed response, restrict server-side request destinations, redirect handling, and approved network egress to prevent SSRF., produces the measurable result requests targeting metadata or private IP ranges fail while sanctioned external URLs still work. The observation tests a relevant security outcome rather than an unrelated control or reporting activity.

Incorrect Answers

A:

In the problem “A service retrieves user-provided URLs on its backend and can contact cloud instance metadata addresses”, evidence “The report separately records verified archive activity and any proven external transmission.” is not decisive because it tests “An enterprise creates a security incident report with unverified statements that data was exfiltrated because a large archive file exists”; the requested validation is requests targeting metadata or private IP ranges fail while sanctioned external URLs still work.

C:

The observed “An altered or untrusted tooling package fails the signed provenance gate” validates the separate control “Validate supplier integrity, package signatures, and the software dependency chain before promotion” under “A container image includes untrusted firmware update tooling supplied by a partner without signature or dependency review”. The present decision calls for the different result “Requests targeting metadata or private IP ranges fail while sanctioned external URLs still work” from “Restrict server-side request destinations, redirect handling, and approved network egress to prevent SSRF”.

D:

Observation “A controlled reflected-XSS payload is displayed harmlessly without executing in the browser.” would be useful for “A customer portal's search field reflects script content into user responses and the team proposes only a perimeter firewall rule”, but it does not prove that the response to “A service retrieves user-provided URLs on its backend and can contact cloud instance metadata addresses” achieved the relevant outcome, requests targeting metadata or private IP ranges fail while sanctioned external URLs still work.

Question 511

An enterprise security review reports: A software updater downloads a signed package but does not verify that the signature belongs to a trusted release authority. Which response best addresses the underlying security issue?

  1. Restrict server-side request destinations, redirect handling, and approved network egress to prevent SSRF.
  2. Retain necessary request attributes and rule provenance under approved data minimization limits.
  3. Adopt cryptographically secure session-token generation with bounded lifetime and revocation support.
  4. Pin approved signing trust chains and verify artifact provenance and revocation before installation.

Explanation

Correct Answer

D:

Pin approved signing trust chains and verify artifact provenance and revocation before installation. The acceptance check is that a package signed by an unknown authority is rejected even if the signature is mathematically valid. That outcome checks whether the requirement is met, rather than treating completion of a security task as proof of effectiveness.

Incorrect Answers

A:

The option “Restrict server-side request destinations, redirect handling, and approved network egress to prevent SSRF.” assumes an issue like “A service retrieves user-provided URLs on its backend and can contact cloud instance metadata addresses”, not the specific observed weakness “A software updater downloads a signed package but does not verify that the signature belongs to a trusted release authority”; it leaves the evidence goal unmet: a package signed by an unknown authority is rejected even if the signature is mathematically valid.

B:

For “A web security monitoring system forwards only summarized WAF scores and discards matched rule identifiers and request context”, “Retain necessary request attributes and rule provenance under approved data minimization limits” has a clear validation signal: “Analysts can replay a sample detection decision from the stored rule and sanitized request metadata”. But “A software updater downloads a signed package but does not verify that the signature belongs to a trusted release authority” calls for “Pin approved signing trust chains and verify artifact provenance and revocation before installation”; the other intervention cannot demonstrate “A package signed by an unknown authority is rejected even if the signature is mathematically valid”.

C:

“Adopt cryptographically secure session-token generation with bounded lifetime and revocation support” addresses “A customer service platform uses predictable session tokens based on timestamps and user IDs”, with “Statistical tests show unpredictable tokens and stolen sessions can be invalidated promptly” as its acceptance test. Here the specific problem is “A software updater downloads a signed package but does not verify that the signature belongs to a trusted release authority”: it requires “Pin approved signing trust chains and verify artifact provenance and revocation before installation” and evidence that “A package signed by an unknown authority is rejected even if the signature is mathematically valid”.

Question 512

Which test or documented result best shows that the underlying issue was properly addressed? A software updater downloads a signed package but does not verify that the signature belongs to a trusted release authority.

  1. A package signed by an unknown authority is rejected even if the signature is mathematically valid.
  2. The report separately records verified archive activity and any proven external transmission.
  3. An altered or untrusted tooling package fails the signed provenance gate.
  4. A cross-site request lacking a valid token fails without altering payment settings.

Explanation

Correct Answer

A:

This is direct test evidence of the control objective in the case: A software updater downloads a signed package but does not verify that the signature belongs to a trusted release authority. It demonstrates that the proposed response, pin approved signing trust chains and verify artifact provenance and revocation before installation., produces the measurable result a package signed by an unknown authority is rejected even if the signature is mathematically valid. The observation tests a relevant security outcome rather than an unrelated control or reporting activity.

Incorrect Answers

B:

Observation “The report separately records verified archive activity and any proven external transmission.” would be useful for “An enterprise creates a security incident report with unverified statements that data was exfiltrated because a large archive file exists”, but it does not prove that the response to “A software updater downloads a signed package but does not verify that the signature belongs to a trusted release authority” achieved the relevant outcome, a package signed by an unknown authority is rejected even if the signature is mathematically valid.

C:

In the problem “A software updater downloads a signed package but does not verify that the signature belongs to a trusted release authority”, evidence “An altered or untrusted tooling package fails the signed provenance gate.” is not decisive because it tests “A container image includes untrusted firmware update tooling supplied by a partner without signature or dependency review”; the requested validation is a package signed by an unknown authority is rejected even if the signature is mathematically valid.

D:

For the issue “A software updater downloads a signed package but does not verify that the signature belongs to a trusted release authority”, measuring “A cross-site request lacking a valid token fails without altering payment settings.” would validate “A cross-origin site can submit authenticated payment settings changes because the application does not validate request intent” instead of the required control effect: a package signed by an unknown authority is rejected even if the signature is mathematically valid.

Question 513

During an authorized security assessment, the team discovers: A multi-threaded account service checks account funds, then updates balance later without atomic protection, allowing double spending. What should be prioritized?

  1. Join application request identity to backend data access records with privacy-appropriate correlation identifiers.
  2. Use parameterized queries and restrict the database identity to needed operations.
  3. Use transactional atomic operations or appropriate locking around the shared balance invariant.
  4. Apply context-specific output encoding in the application with defensive validation and regression tests.

Explanation

Correct Answer

C:

Use transactional atomic operations or appropriate locking around the shared balance invariant. The acceptance check is that concurrent test withdrawals cannot exceed the authorized balance. That outcome checks whether the requirement is met, rather than treating completion of a security task as proof of effectiveness.

Incorrect Answers

A:

This choice would be evaluated by “A suspicious query can be attributed to the initiating authorized or unauthorized user session” when dealing with “A SOC analyst needs to investigate suspicious database usage but logs list only generic application service IDs”. It does not deliver the distinct control “Use transactional atomic operations or appropriate locking around the shared balance invariant” needed for “A multi-threaded account service checks account funds, then updates balance later without atomic protection, allowing double spending”.

B:

Although “Use parameterized queries and restrict the database identity to needed operations.” is a reasonable response to “A web backend concatenates an untrusted address parameter into a SQL query that reads personal customer records”, this case instead concerns “A multi-threaded account service checks account funds, then updates balance later without atomic protection, allowing double spending”, for which success requires that concurrent test withdrawals cannot exceed the authorized balance.

D:

The proposal “Apply context-specific output encoding in the application with defensive validation and regression tests” manages “A customer portal's search field reflects script content into user responses and the team proposes only a perimeter firewall rule”, where successful operation means “A controlled reflected-XSS payload is displayed harmlessly without executing in the browser”. The present issue instead needs “Use transactional atomic operations or appropriate locking around the shared balance invariant” to establish “Concurrent test withdrawals cannot exceed the authorized balance”.

Question 514

A follow-up assurance review considers this problem: A multi-threaded account service checks account funds, then updates balance later without atomic protection, allowing double spending. Which outcome best shows the intended improvement?

  1. Concurrent test withdrawals cannot exceed the authorized balance.
  2. A crafted destination override cannot redirect another user's reset token.
  3. A suspicious query can be attributed to the initiating authorized or unauthorized user session.
  4. Traversal tests cannot read configuration files outside the approved root.

Explanation

Correct Answer

A:

This is direct test evidence of the control objective in the case: A multi-threaded account service checks account funds, then updates balance later without atomic protection, allowing double spending. It demonstrates that the proposed response, use transactional atomic operations or appropriate locking around the shared balance invariant., produces the measurable result concurrent test withdrawals cannot exceed the authorized balance. The observation tests a relevant security outcome rather than an unrelated control or reporting activity.

Incorrect Answers

B:

This measurement is tied to “A password-reset endpoint trusts the caller's requested username and issues reset tokens to an attacker-controlled address” and the control “Verify recovery destination ownership through authoritative account records and controlled out-of-band confirmation”. It cannot confirm that “Use transactional atomic operations or appropriate locking around the shared balance invariant” has achieved “Concurrent test withdrawals cannot exceed the authorized balance” for the issue in this question.

C:

“A suspicious query can be attributed to the initiating authorized or unauthorized user session” would verify “Join application request identity to backend data access records with privacy-appropriate correlation identifiers” for the situation “A SOC analyst needs to investigate suspicious database usage but logs list only generic application service IDs”. Here the required evidence is “Concurrent test withdrawals cannot exceed the authorized balance” after “Use transactional atomic operations or appropriate locking around the shared balance invariant”; the alternative observation does not test that control.

D:

Observation “Traversal tests cannot read configuration files outside the approved root.” would be useful for “A retail service receives requests whose raw path components traverse outside the intended static file directory”, but it does not prove that the response to “A multi-threaded account service checks account funds, then updates balance later without atomic protection, allowing double spending” achieved the relevant outcome, concurrent test withdrawals cannot exceed the authorized balance.

Question 515

A cross-functional security group is considering a decision: An enterprise continues deploying a database library after its vendor ends security support and no maintenance fork exists. What is the most appropriate control or investigation response?

  1. Validate supplier integrity, package signatures, and the software dependency chain before promotion.
  2. Include sensitivity and effective access evidence when interpreting cloud posture findings.
  3. Disable obsolete cryptographic suites and require a supported authenticated TLS profile.
  4. Replace the library with a supported alternative and track residual transition risk until the rollout.

Explanation

Correct Answer

D:

Replace the library with a supported alternative and track residual transition risk until the rollout. The acceptance check is that new builds reject the unsupported package and the supported replacement passes functional and security tests. That outcome checks whether the requirement is met, rather than treating completion of a security task as proof of effectiveness.

Incorrect Answers

A:

The option “Validate supplier integrity, package signatures, and the software dependency chain before promotion.” assumes an issue like “A container image includes untrusted firmware update tooling supplied by a partner without signature or dependency review”, not the specific observed weakness “An enterprise continues deploying a database library after its vendor ends security support and no maintenance fork exists”; it leaves the evidence goal unmet: new builds reject the unsupported package and the supported replacement passes functional and security tests.

B:

“Priority decisions separately document low-sensitivity public objects and exposed personnel records” would be meaningful evidence for “Include sensitivity and effective access evidence when interpreting cloud posture findings” in the case “An enterprise faces this security constraint: A cloud posture tool emits a risk score for publicly accessible data storage but does not distinguish public marketing content from HR records”. It does not replace “Replace the library with a supported alternative and track residual transition risk until the rollout” to correct “A cross-functional security group is considering a decision: An enterprise continues deploying a database library after its vendor ends security support and no maintenance fork exists” and verify “New builds reject the unsupported package and the supported replacement passes functional and security tests”.

C:

For “A legacy service uses a weak block cipher and accepts a downgrade when a modern client negotiates connections”, “Disable obsolete cryptographic suites and require a supported authenticated TLS profile” has a clear validation signal: “Client tests reject weak negotiation and approved modern clients connect successfully”. But “A cross-functional security group is considering a decision: An enterprise continues deploying a database library after its vendor ends security support and no maintenance fork exists” calls for “Replace the library with a supported alternative and track residual transition risk until the rollout”; the other intervention cannot demonstrate “New builds reject the unsupported package and the supported replacement passes functional and security tests”.

Question 516

After the team responds to the following issue, which finding best demonstrates the desired security outcome? An enterprise continues deploying a database library after its vendor ends security support and no maintenance fork exists.

  1. Newly created test containers appear in monitoring and security events within the ingestion SLA.
  2. A package signed by an unknown authority is rejected even if the signature is mathematically valid.
  3. Client tests reject weak negotiation and approved modern clients connect successfully.
  4. New builds reject the unsupported package and the supported replacement passes functional and security tests.

Explanation

Correct Answer

D:

This is direct test evidence of the control objective in the case: An enterprise continues deploying a database library after its vendor ends security support and no maintenance fork exists. It demonstrates that the proposed response, replace the library with a supported alternative and track residual transition risk until the rollout., produces the measurable result new builds reject the unsupported package and the supported replacement passes functional and security tests. The observation tests a relevant security outcome rather than an unrelated control or reporting activity.

Incorrect Answers

A:

This measurement is tied to “An application changes from fixed IP infrastructure to autoscaling containers, but monitoring rules are pinned to the former host list” and the control “Adopt workload identity and dynamic inventory enrichment for telemetry and alert coverage”. It cannot confirm that “Replace the library with a supported alternative and track residual transition risk until the rollout” has achieved “New builds reject the unsupported package and the supported replacement passes functional and security tests” for the issue in this question.

B:

Observation “A package signed by an unknown authority is rejected even if the signature is mathematically valid.” would be useful for “A software updater downloads a signed package but does not verify that the signature belongs to a trusted release authority”, but it does not prove that the response to “An enterprise continues deploying a database library after its vendor ends security support and no maintenance fork exists” achieved the relevant outcome, new builds reject the unsupported package and the supported replacement passes functional and security tests.

C:

In the problem “An enterprise continues deploying a database library after its vendor ends security support and no maintenance fork exists”, evidence “Client tests reject weak negotiation and approved modern clients connect successfully.” is not decisive because it tests “A legacy service uses a weak block cipher and accepts a downgrade when a modern client negotiates connections”; the requested validation is new builds reject the unsupported package and the supported replacement passes functional and security tests.

Question 517

In a control effectiveness review, the issue is: A password-reset endpoint trusts the caller's requested username and issues reset tokens to an attacker-controlled address. Which intervention is most defensible for the situation?

  1. Verify recovery destination ownership through authoritative account records and controlled out-of-band confirmation.
  2. Correlate DLP with data classification, user approval, destination, and observed file activity.
  3. Use safe data formats and strict allowed-type validation in the import path.
  4. Pin approved signing trust chains and verify artifact provenance and revocation before installation.

Explanation

Correct Answer

A:

Verify recovery destination ownership through authoritative account records and controlled out-of-band confirmation. The acceptance check is that a crafted destination override cannot redirect another user's reset token. That outcome checks whether the requirement is met, rather than treating completion of a security task as proof of effectiveness.

Incorrect Answers

B:

“An investigation distinguishes an authorized export from a policy-violating personal destination” would be meaningful evidence for “Correlate DLP with data classification, user approval, destination, and observed file activity” in the case “A network SOC receives many DLP findings but cannot tell whether uploads were approved data exports or accidental personal file shares”. It does not replace “Verify recovery destination ownership through authoritative account records and controlled out-of-band confirmation” to correct “A password-reset endpoint trusts the caller's requested username and issues reset tokens to an attacker-controlled address” and verify “A crafted destination override cannot redirect another user's reset token”.

C:

The proposal “Use safe data formats and strict allowed-type validation in the import path” manages “A cross-functional security group is considering a decision: An internal build runner loads untrusted serialized objects from a developer-submitted artifact with executable type handlers enabled”, where successful operation means “A crafted object cannot execute unintended code during a controlled build”. The present issue instead needs “Verify recovery destination ownership through authoritative account records and controlled out-of-band confirmation” to establish “A crafted destination override cannot redirect another user's reset token”.

D:

For “A software updater downloads a signed package but does not verify that the signature belongs to a trusted release authority”, “Pin approved signing trust chains and verify artifact provenance and revocation before installation” has a clear validation signal: “A package signed by an unknown authority is rejected even if the signature is mathematically valid”. But “A password-reset endpoint trusts the caller's requested username and issues reset tokens to an attacker-controlled address” calls for “Verify recovery destination ownership through authoritative account records and controlled out-of-band confirmation”; the other intervention cannot demonstrate “A crafted destination override cannot redirect another user's reset token”.

Question 518

Which test or documented result best shows that the underlying issue was properly addressed? A password-reset endpoint trusts the caller's requested username and issues reset tokens to an attacker-controlled address.

  1. A crafted destination override cannot redirect another user's reset token.
  2. A crafted malformed login produces an ingestion error and cannot be mistaken for an authenticated success.
  3. A crafted object cannot execute unintended code during a controlled build.
  4. Concurrent test withdrawals cannot exceed the authorized balance.

Explanation

Correct Answer

A:

This is direct test evidence of the control objective in the case: A password-reset endpoint trusts the caller's requested username and issues reset tokens to an attacker-controlled address. It demonstrates that the proposed response, verify recovery destination ownership through authoritative account records and controlled out-of-band confirmation., produces the measurable result a crafted destination override cannot redirect another user's reset token. The observation tests a relevant security outcome rather than an unrelated control or reporting activity.

Incorrect Answers

B:

“A crafted malformed login produces an ingestion error and cannot be mistaken for an authenticated success” would verify “Validate schema and reject malformed events with explicit parser health errors” for the situation “A cross-functional security group is considering a decision: A developer uses a wildcard parser that converts missing security log fields to zero values, making failed imports look like successful logins”. Here the required evidence is “A crafted destination override cannot redirect another user's reset token” after “Verify recovery destination ownership through authoritative account records and controlled out-of-band confirmation”; the alternative observation does not test that control.

C:

In “A cross-functional security group is considering a decision: An internal build runner loads untrusted serialized objects from a developer-submitted artifact with executable type handlers enabled”, “A crafted object cannot execute unintended code during a controlled build” is the acceptance signal for “Use safe data formats and strict allowed-type validation in the import path”. The required proof for the selected response “Verify recovery destination ownership through authoritative account records and controlled out-of-band confirmation” is “A crafted destination override cannot redirect another user's reset token”.

D:

The observed “Concurrent test withdrawals cannot exceed the authorized balance” validates the separate control “Use transactional atomic operations or appropriate locking around the shared balance invariant” under “A multi-threaded account service checks account funds, then updates balance later without atomic protection, allowing double spending”. The present decision calls for the different result “A crafted destination override cannot redirect another user's reset token” from “Verify recovery destination ownership through authoritative account records and controlled out-of-band confirmation”.

Question 519

A security architect is asked to resolve this problem: A customer service platform uses predictable session tokens based on timestamps and user IDs. Select the MOST directly relevant action.

  1. Canonicalize paths and constrain server file access to an authorized directory after validation.
  2. Pin approved signing trust chains and verify artifact provenance and revocation before installation.
  3. Adopt workload identity and dynamic inventory enrichment for telemetry and alert coverage.
  4. Adopt cryptographically secure session-token generation with bounded lifetime and revocation support.

Explanation

Correct Answer

D:

Adopt cryptographically secure session-token generation with bounded lifetime and revocation support. The acceptance check is that statistical tests show unpredictable tokens and stolen sessions can be invalidated promptly. That outcome checks whether the requirement is met, rather than treating completion of a security task as proof of effectiveness.

Incorrect Answers

A:

The option “Canonicalize paths and constrain server file access to an authorized directory after validation.” assumes an issue like “A retail service receives requests whose raw path components traverse outside the intended static file directory”, not the specific observed weakness “A customer service platform uses predictable session tokens based on timestamps and user IDs”; it leaves the evidence goal unmet: statistical tests show unpredictable tokens and stolen sessions can be invalidated promptly.

B:

The proposal “Pin approved signing trust chains and verify artifact provenance and revocation before installation” manages “A software updater downloads a signed package but does not verify that the signature belongs to a trusted release authority”, where successful operation means “A package signed by an unknown authority is rejected even if the signature is mathematically valid”. The present issue instead needs “Adopt cryptographically secure session-token generation with bounded lifetime and revocation support” to establish “Statistical tests show unpredictable tokens and stolen sessions can be invalidated promptly”.

C:

For “An application changes from fixed IP infrastructure to autoscaling containers, but monitoring rules are pinned to the former host list”, “Adopt workload identity and dynamic inventory enrichment for telemetry and alert coverage” has a clear validation signal: “Newly created test containers appear in monitoring and security events within the ingestion SLA”. But “A customer service platform uses predictable session tokens based on timestamps and user IDs” calls for “Adopt cryptographically secure session-token generation with bounded lifetime and revocation support”; the other intervention cannot demonstrate “Statistical tests show unpredictable tokens and stolen sessions can be invalidated promptly”.

Question 520

A follow-up assurance review considers this problem: A customer service platform uses predictable session tokens based on timestamps and user IDs. Which outcome best shows the intended improvement?

  1. A completed hunt records hypothesis, supporting or contradictory data, and implemented detection improvements.
  2. Network tests show obsolete ports closed while authorized administration remains available.
  3. Statistical tests show unpredictable tokens and stolen sessions can be invalidated promptly.
  4. Client tests reject weak negotiation and approved modern clients connect successfully.

Explanation

Correct Answer

C:

This is direct test evidence of the control objective in the case: A customer service platform uses predictable session tokens based on timestamps and user IDs. It demonstrates that the proposed response, adopt cryptographically secure session-token generation with bounded lifetime and revocation support., produces the measurable result statistical tests show unpredictable tokens and stolen sessions can be invalidated promptly. The observation tests a relevant security outcome rather than an unrelated control or reporting activity.

Incorrect Answers

A:

This measurement is tied to “A threat-hunting program reports the number of searches performed rather than whether meaningful adversary hypotheses were tested” and the control “Define evidence-based hunt outcomes and detection engineering feedback loops”. It cannot confirm that “Adopt cryptographically secure session-token generation with bounded lifetime and revocation support” has achieved “Statistical tests show unpredictable tokens and stolen sessions can be invalidated promptly” for the issue in this question.

B:

Observation “Network tests show obsolete ports closed while authorized administration remains available.” would be useful for “A network appliance has default administrative services enabled despite the organization never using them”, but it does not prove that the response to “A customer service platform uses predictable session tokens based on timestamps and user IDs” achieved the relevant outcome, statistical tests show unpredictable tokens and stolen sessions can be invalidated promptly.

D:

For the issue “A customer service platform uses predictable session tokens based on timestamps and user IDs”, measuring “Client tests reject weak negotiation and approved modern clients connect successfully.” would validate “A legacy service uses a weak block cipher and accepts a downgrade when a modern client negotiates connections” instead of the required control effect: statistical tests show unpredictable tokens and stolen sessions can be invalidated promptly.