On this page

RNX code library · 7 controls

RNX-700Resilience, Adversarial Defense & Future Compatibility

Read what these checks mean, why they matter and which evidence they concern. Each ID below keeps its canonical place in the RNX hierarchy.

These are free learning summaries, not full technical tests or an industry certification. Specialized checks apply only to the relevant feature. Experimental labels are preserved from the canonical source.

RNX-710 · Adversarial Entity Defense

RNX-711 · Synthetic Citation Poisoning Detection

RNX-711.T001

Suspicious citation investigation

Experimental — not an established ranking requirement

Investigate suspicious citations without assuming manipulation.

Why it matters. Similar wording alone cannot prove a coordinated attack.

Illustration — fictional. Similar words alone called coordinated attack.

When it applies. Inspected target: anomaly finding. Use this check only when that object or feature exists in the review scope; provider-specific checks require that provider's integration. Missing evidence remains unknown.

Canonical parentage

RNX-700 · Resilience, Adversarial Defense & Future Compatibility RNX-710 · Adversarial Entity Defense RNX-711 · Synthetic Citation Poisoning Detection

RNX-712 · Competitor-Agent Spoofing Defense

RNX-712.T001

Protected caller identity verification

core readiness

For protected services, verify the caller rather than trusting its copied public identity.

Why it matters. A familiar name or agent card is not authentication.

Illustration — fictional. A caller's copied public agent name is trusted instead of checking authorization.

When it applies. Inspected target: caller fixture. Use this check only when that object or feature exists in the review scope; provider-specific checks require that provider's integration. Missing evidence remains unknown.

Canonical parentage

RNX-700 · Resilience, Adversarial Defense & Future Compatibility RNX-710 · Adversarial Entity Defense RNX-712 · Competitor-Agent Spoofing Defense

RNX-715 · Retrieval Field Quarantine Protocol

RNX-715.T001

Compromised-source containment

Experimental — not an established ranking requirement

Contain compromised content in owned sources and saved copies.

Why it matters. Correcting the live page alone may leave a harmful older version in use.

Illustration — fictional. The website is repaired, but an owned answer system still retrieves the compromised saved copy.

When it applies. Inspected target: containment fixture. Use this check only when that object or feature exists in the review scope; provider-specific checks require that provider's integration. Missing evidence remains unknown.

Canonical parentage

RNX-700 · Resilience, Adversarial Defense & Future Compatibility RNX-710 · Adversarial Entity Defense RNX-715 · Retrieval Field Quarantine Protocol

RNX-720 · Small Model & Local Retrieval Compatibility

RNX-721 · Ultra-Compact Entity Summary

RNX-721.T001

Budgeted identity-summary fidelity

Experimental — not an established ranking requirement

For limited-length identity summaries, preserve the facts that distinguish the business.

Why it matters. A shorter summary should not merge two same-name firms.

Illustration — fictional. Summary merges same-named firms.

When it applies. Inspected target: summary fidelity. Use this check only when that object or feature exists in the review scope; provider-specific checks require that provider's integration. Missing evidence remains unknown.

Canonical parentage

RNX-700 · Resilience, Adversarial Defense & Future Compatibility RNX-720 · Small Model & Local Retrieval Compatibility RNX-721 · Ultra-Compact Entity Summary

RNX-724 · Context-Minimal Trust Bundle

RNX-724.T001

Supported trust-bundle validation

Experimental — not an established ranking requirement

For supported bundles of trust evidence, check format and evidence links.

Why it matters. A readable bundle is not a certification of trustworthiness.

Illustration — fictional. A readable evidence bundle points to expired proof about a different business.

When it applies. Inspected target: trust bundle. Use this check only when that object or feature exists in the review scope; provider-specific checks require that provider's integration. Missing evidence remains unknown.

Canonical parentage

RNX-700 · Resilience, Adversarial Defense & Future Compatibility RNX-720 · Small Model & Local Retrieval Compatibility RNX-724 · Context-Minimal Trust Bundle

RNX-730 · Multimodal Authority Standard

RNX-735 · Multimodal Provenance Watermarking

RNX-735.T001

Content credential verification

core readiness

For content credentials, report what verification can establish and what is absent.

Why it matters. An unsigned image is not automatically fake, and a signed one is not automatically true.

Illustration — fictional. Unsigned image labeled fake.

When it applies. Inspected target: credential fixture. Use this check only when that object or feature exists in the review scope; provider-specific checks require that provider's integration. Missing evidence remains unknown.

Canonical parentage

RNX-700 · Resilience, Adversarial Defense & Future Compatibility RNX-730 · Multimodal Authority Standard RNX-735 · Multimodal Provenance Watermarking

RNX-791 · RNX Compliance Scorecard

RNX-791.T001

Scoped RNX assessment claim

Experimental — not an established ranking requirement

Describe an RNX assessment by its actual version, scope and reviewer.

Why it matters. An RNX review must not be called Google certification or independent approval without evidence.

Illustration — fictional. RNX review called Google certification.

When it applies. Inspected target: assessment claim. Use this check only when that object or feature exists in the review scope; provider-specific checks require that provider's integration. Missing evidence remains unknown.

Canonical parentage

RNX-700 · Resilience, Adversarial Defense & Future Compatibility RNX-791 · RNX Compliance Scorecard

Related learning