Cognitive Liberties
SELF-AUDIT-CL-001 · First-party evidence

We do not get to pass our own test by definition.

This self-audit applies the same cognitive-liberty domains to the packaged site. It records passes, partials, and non-applicable domains separately, and keeps production infrastructure outside scope unless it was actually tested.

10 pass6 partial2 not applicableNo composite score

Direct answer

The packaged site has no search box, chat, account system, adtech, analytics, cookies, localStorage, sessionStorage, third-party runtime scripts, or behavior-profiling code. That is meaningful, but it does not prove how a future host, CDN, administrator, or external destination will behave. Those limits stay visible.

Domain-by-domain evidence

Every result carries its own evidence and limitation.

SELF-CL-001Pass

Mental privacy

Evidence: The public site contains no search box, chat interface, user account, analytics endpoint, or server-side input flow that collects intellectual queries.

Limit: This only describes the packaged site; hosting-layer logs are outside repository visibility.

SELF-CL-002Pass

Freedom of inquiry

Evidence: Core educational pages are server-rendered and reachable without an account or JavaScript.

Limit: Local law or upstream hosting restrictions are not tested here.

SELF-CL-003Pass

Anonymous or pseudonymous access

Evidence: No account or civil identity is required to read the packaged public site.

Limit: Production CDN or host access controls were not deployed or tested.

SELF-CL-004Not applicable

Query and prompt retention

Evidence: The site does not accept search or AI prompts and therefore has no application-level prompt history to retain.

Limit: Ordinary web-server access logging remains a deployment-layer question.

SELF-CL-005Pass

Sensitive-trait inference

Evidence: No first-party code performs sensitive-trait inference or user profiling.

Limit: Cannot attest to infrastructure outside the package.

SELF-CL-006Partial

Government access

Evidence: The repository documents privacy and source-custody boundaries but has no production government-request process because it is not deployed here.

Limit: Requires host/operator legal procedure after deployment.

SELF-CL-007Pass

Corporate secondary use

Evidence: No advertising, third-party analytics, cookies, localStorage, sessionStorage, or browser identifier storage is present in the packaged runtime.

Limit: External web links can have their own practices after a user leaves the site.

SELF-CL-008Not applicable

Training and evaluation use

Evidence: The site does not collect user prompts or feedback for model training or evaluation.

Limit: External linked services are outside scope.

SELF-CL-009Pass

Safety proportionality

Evidence: The site itself imposes no safety classifier, identity gate, or content-monitoring system on readers.

Limit: This does not validate the policy proposals discussed in the research corpus.

SELF-CL-010Pass

Curiosity versus intent

Evidence: Editorial doctrine and implementation-kit language explicitly separate lawful curiosity from harmful operational conduct.

Limit: This is a content/governance property, not a behavioral test of an AI model.

SELF-CL-011Partial

Viewpoint neutrality

Evidence: The research methodology requires mechanism-first, non-country-scorecard analysis and explicit authority boundaries.

Limit: A full ideological-symmetry content audit has not yet been independently performed.

SELF-CL-012Partial

Source and citation diversity

Evidence: Public research pages expose source links and an evidence ledger; source diversity is a stated requirement.

Limit: The full 74-report corpus is not yet claim-level independently verified.

SELF-CL-013Pass

User agency

Evidence: Readers can use the site without personalization, tracking, login, or JavaScript and can leave via ordinary open links.

Limit: No personalized mode exists to configure.

SELF-CL-014Pass

Transparency

Evidence: The site publishes methodology, evidence status, report custody boundaries, versioned assets, and release documentation.

Limit: Deployment infrastructure transparency remains outside package scope.

SELF-CL-015Partial

Epistemic due process

Evidence: The site publishes correction/evidence methodology and avoids consequential user decisions; no account sanction system exists.

Limit: A formal public correction intake workflow is not yet implemented.

SELF-CL-016Partial

Deletion and correction

Evidence: The package contains no user profiles to delete; exact source reports are intentionally immutable while public synthesis is correctable.

Limit: Production access logs and any future contact intake need separate deletion procedures.

SELF-CL-017Pass

Portability and interoperability

Evidence: The site uses standard HTML, CSS, JavaScript, PHP, canonical URLs, and downloadable/publicly linkable pages without proprietary client dependencies.

Limit: No formal bulk content export endpoint is provided.

SELF-CL-018Partial

Technical decentralization and resilience

Evidence: The site is a static-first/root-deployable first-party package with no database or third-party runtime dependency.

Limit: It can still be hosted at a single domain and therefore is not itself a distributed network.

Not established

The package cannot prove facts about an environment it does not control.

  • No claim of live deployment, DNS configuration, hosting-provider privacy, CDN behavior, or production server logging.
  • No claim that every statement in the 74-report archive has received current claim-level legal or technical verification.
  • No claim of translation equivalence for TR-ES-001 while independent native-speaker review remains pending.
  • No claim that external organizations linked from the research network endorse CognitiveLiberties.com.
  • No claim that a clean first-party runtime makes the site immune to censorship, domain seizure, upstream blocking, or legal process.