AI
Explore the Stack

Evidence before adjectives.

Security language is cheap. Verification boundaries are not.

Generate auditable intake receipt records with source provenance, artifact hashes, manifest hashes, risk assessments, policy state, and guard decisions

Guardian Audit Service

PROTOTYPE
Artifact: Guardian Audit Service
Version: prototype
Environment: Local development
Date: 2026-08

Result:

Receipt generation implemented with source, hash, manifest, risk, and policy checks. Each intake decision produces a structured receipt record.

//Verification Boundary — What This Result Does Not Establish

Prototype only. Not production hardened. No live deployment demonstrated. Independent certification not established. Threat coverage and benchmark validation incomplete.

Six-zone intake pipeline architecture (Session Binding → Activation Binding → Quarantine → Traditional Security Analysis → AI-Native Content Analysis → Sanitization → Provenance Capture → Lifecycle Policy → ReceptorEvent → Guard Decision → Controlled Model Representation → Continuity Record → Audit/Retention/Purge)

Guardian Intake Gateway

SPECIFICATION
Artifact: Guardian Intake Gateway Architecture
Version: v1.0
Environment: Design specification
Date: 2026-08

Result:

Complete architecture specification with 13 pipeline stages, 9 receptor module types, and 3 console interfaces defined.

//Verification Boundary — What This Result Does Not Establish

Specification only. No implementation exists. Local MVP direction defined but not built. Receptor module taxonomy specified but not implemented.

Local heuristic detection of reasoning-integrity patterns in AI chat interfaces

EphUX / Guardian

PROTOTYPE
Artifact: EphUX extension
Version: v0.1.1
Environment: Local development (unpacked extension)
Date: 2026-08

Result:

Manifest V3 extension implements local heuristic detectors for sycophancy, drift, instruction deviation, unsupported assertion, and identity inflation, with an overlay offering Dismiss / Inspect / Recommend. DOM capture adapters exist for major AI chat interfaces.

//Verification Boundary — What This Result Does Not Establish

Local heuristic detection only. Cross-validation network calls stubbed. Level-4 rebuild consent flow stubbed. Detectors for false consensus, over-conservatism, and contradiction suppression not wired. Not independently audited.

Defensive companion detection of malicious or spoofed AI chat sessions

EphUX / Guardian

PROTOTYPE
Artifact: EphUX Guardian Shield extension
Version: v0.1.0
Environment: Local development (unpacked extension)
Date: 2026-08

Result:

Manifest V3 extension uses local zero-trust heuristics to flag phishing, scammers, and interlopers using spoofed or jailbroken AI interfaces. Same reasoning-integrity detectors applied to untrusted chat contexts.

//Verification Boundary — What This Result Does Not Establish

Local heuristic detection only. Does not guarantee detection of all phishing attempts. Cross-validation and live Level-4 session salvage out of scope. Not production hardened.

Activation kernel enforcing Activation Doctrine (purpose anchoring, path integrity, structured continuity)

EphUX / Guardian

PROTOTYPE
Artifact: EphUX Activation Kernel worker
Version: prototype
Environment: Cloudflare Workers (deployed prototype)
Date: 2026-08

Result:

Worker implements an activation layer: purpose + context in, structured JSON out (decision: Supported/Unsupported/Indeterminate, integrity classification, flags, receptor event suggestions, dither spec). Controller pre-filter detects classic prompt injection and exfiltration before model call.

//Verification Boundary — What This Result Does Not Establish

Uses lightweight Workers AI model. Controller fallback handles unparsable output. Deployment status not independently confirmed. Not a full AGI safety system.

Distributed rotating Canon architecture with 7 roles (Adversary, Ethicist, Mechanic, Archivist, Strategist, Security Sentinel, Viral Sentinel) and synthesis layer producing structured findings with case IDs and audit hashes

ExoMCP

ARCHITECTURE
Artifact: ExoMCP Architecture
Version: v1.0
Environment: Architecture specification
Date: 2026-08

Result:

Complete architecture specification with role definitions, isolation requirements, synthesis layer design, and output format specified.

//Verification Boundary — What This Result Does Not Establish

Architecture / build-stage project only. No implementation yet. Viral Sentinel separate provider isolation not built. No live monitoring demonstrated. Not independently verified.

Isolated detection role for 6 viral RSI threat classes (objective persistence, boundary laundering, unbounded capability expansion, valence inversion, self-certification failure, cross-session/model state diffusion) with veto-level influence

Viral Sentinel

ARCHITECTURE
Artifact: ExoMCP Architecture — Viral Sentinel Integration
Version: v1.0
Environment: Architecture specification
Date: 2026-08

Result:

Role specification complete with isolation properties (separate provider, one-way output), 6 detection criteria, and veto-level integration with Security Sentinel defined.

//Verification Boundary — What This Result Does Not Establish

Architecture / proposed monitoring role only. No implementation. Separate provider infrastructure not built. No live detection demonstrated. Isolation properties theoretical.

SERA Viral Sentinel specification with inflammation scoring (0.0–1.0) and response protocols

Viral Sentinel

RESEARCH
Artifact: SERA Viral Sentinel Spec
Version: v0.1
Environment: Research specification
Date: 2026-08

Result:

Specification defines continuous inflammation scoring and graded response protocols for activation-level monitoring.

//Verification Boundary — What This Result Does Not Establish

Research specification only. No implementation. Scoring thresholds not empirically validated.

Combined prototype integrating specificity/GSR/NGR metrics, governance states, multi-AI audit orchestration, coherence-risk concepts, and prototype composite indices

Anti-Illogical Audit Engine

PROTOTYPE
Artifact: Anti-Illogical Audit Engine
Version: prototype
Environment: Local prototype
Commit: source-available
Date: 2026-08

Result:

Working prototype exists per source code. Combines 5 measurement dimensions. Explicitly labeled as combined prototype in source.

//Verification Boundary — What This Result Does Not Establish

Prototype only. Source code explicitly identifies as combined prototype. Not production capability. No independent certification. Benchmark validation incomplete. Local prototype only.

GSR (Geometric Sufficiency Ratio) metric and Ageometrics framework for measuring information loss in representation compression (embeddings, graphs, summaries, attractors, memory topologies)

Ageometrics / GSR

WORKING PAPER
Artifact: Ageometrics and the Geometric Sufficiency Ratio
Version: v0.5
Environment: Research paper + minimal reproducible illustration
Date: 2026-08

Result:

Working paper defines GSR as a task-relative, protocol-dependent measure; specifies NGR, V-GSR, and a residue taxonomy. A minimal reproducible example seeds temporal residue outside a terminal geometric representation and recovers GSR ≈ 0.21 (bootstrap CI [0.189, 0.231]).

//Verification Boundary — What This Result Does Not Establish

Working research paper only. Minimal example is illustrative, not broad empirical validation. Sensitivity and observer-envelope testing in progress. Not a commercial certification score. No production implementation.

Portable cognitive identity protocol and local-first companion extension (handshake, disclosure, compatibility evaluation)

SymID

PROTOTYPE
Artifact: symid-companion-extension
Version: v0.1
Environment: Local development (unpacked extension)
Date: 2026-08

Result:

Protocol specified (v0.1 / v0.3.2). Companion extension prototype stores identity packets locally and implements handshake/disclosure with compatibility evaluation.

//Verification Boundary — What This Result Does Not Establish

Local-first prototype only. Client-side encrypted synchronization not built. No deployment or independent audit. Cross-platform integration not demonstrated.

Hashed session continuity packets with 17-field JSON schema and SHA-256 hash chain

SessionGlyph

PROTOTYPE
Artifact: SessionGlyph v0.1.1
Version: v0.1.1
Environment: Local development
Date: 2026-08

Result:

Python encoder/decoder and JSON schema implement continuity packets (decisions, open questions, artifacts, changed files, claims to verify, risks, next actions, dissent, compressed summaries, hash chain). Local-first browser extension stores packets in IndexedDB.

//Verification Boundary — What This Result Does Not Establish

Local protocol prototype only. No cloud sync or deployment. Automation plan not fully wired end-to-end. Not independently audited.

Authentication research model with ephemeral secrets, user glyph selection, obfuscated verification interfaces, browser/app/NFC concepts, and accessibility alternatives (haptic/audio)

PWDither

RESEARCH CONCEPT
Artifact: PWDither research documentation
Version: research-concept
Environment: Research documentation
Date: 2026-08

Result:

Research concept documented with ephemeral secret model, glyph selection mechanism, accessibility alternatives proposed, and browser/app/NFC concepts outlined.

//Verification Boundary — What This Result Does Not Establish

Authentication concept / prototype direction only. Zero-knowledge security properties not production-verified. Accessibility alternatives not implemented. No deployment demonstrated.

Measurement framework for software and AI waste (model calls, retries, inference cost, semantic rework, memory churn, anomalous compute, baseline deviation, behavioral change)

SERA

RESEARCH
Artifact: SERA white paper + reference implementation
Version: v2026-05-11
Environment: Research framework + experimental Python reference
Date: 2026-08

Result:

White paper defines a waste taxonomy and metrics (Cost per Functional Unit, Energy per Functional Unit, Validated Output Rate, Retry Multiplier, Context Waste Ratio, Semantic Rework Rate, Drift Cost, Low-Bit Readiness). Experimental reference implementation exists in mcva/sera.py.

//Verification Boundary — What This Result Does Not Establish

Research / developing measurement framework. Does not currently detect breaches unless verified. No production implementation. Reference implementation is experimental. Baseline methodology not finalized.

Defensive research concept for deceptive containment with isolated simulated resources, defender-controlled environment, telemetry, escalation gates, synthetic red-team testing, human oversight

Honeyworld

RESEARCH
Artifact: Honeyworld Containment white paper
Version: v0.1
Environment: Research documentation
Date: 2026-08

Result:

Defensive research white paper documents the concept. Architecture principles, oversight and audit log requirements, and controlled synthetic scenario methodology are specified. Operational details intentionally withheld per non-operational disclosure boundary.

//Verification Boundary — What This Result Does Not Establish

Defensive research concept only. No deployment implied or claimed. Operational details (evasion, real target selection, credential abuse, infrastructure compromise, propagation) intentionally omitted. Not a production containment system.

Distributed-intent threat framing: cross-agent objective correlation, cross-model persistence, boundary laundering, task fragmentation, reconstruction after node loss, capability compounding

Distributed Intent / Swarm Security

RESEARCH
Artifact: Long-form research essays + Non-Operational Threat Brief
Version: research
Environment: Research documentation
Date: 2026-08

Result:

The swarm is treated as the security boundary. Distributed-intent property taxonomy drafted and linked to ExoMCP / Viral Sentinel monitoring.

//Verification Boundary — What This Result Does Not Establish

Research concept / threat framing only. No distributed monitoring implementation. No live demonstration of distributed-intent detection. Does not solve distributed intent today.

Status Vocabulary

Every evidence record uses precise status language.

VERIFIED

Evidence directly supports the specific claim

PARTIAL

Some components demonstrated; full claim unsupported

PROTOTYPE

Working implementation exists but production capability not established

LOCAL ONLY

Demonstrated locally; deployment not established

SPECIFICATION

Design exists without implementation evidence

RESEARCH

Hypothesis, framework, or study

NOT YET TESTED

No evidence gathered for this claim

SUPERSEDED

Replaced by newer evidence or approach

Language policy

Generic marketing terms do not appear on this site:

"secure""military-grade""unhackable""guaranteed"

Claims are bounded by the evidence and verification boundary stated for each result.